N PS ARCHIVE 1967 MARSHALL, B. EFFECT Of* SLATOtt HLADii: '■■■ WfAllON ON THE PERFORMANCE OF AN AXIAL FLOW COMPRESSOR imUCK CAMERON MARSHALL 111 i,; ■ • EFFECT OF STATOR BLADE ORIENTATION ON THE PERFORMANCE OF AN AXIAL FLOW COMPRESSOR by Bruce Cameron Marshall Lieutenant, United//3tates Navy B. Ch. E., Cornell lAhiversity, 1959 Submitted in partial fulfillment of the requirements for the degree of AERONAUTICAL ENGINEER from the NAVAL POSTGRADUATE SCHOOL September 1967 ABSTRACT A mean streamline analysis of the effect of stator blade orienta- tion on the performance of an axial flow compressor was performed by means of a computer program. Measurements were made on a 3-stage axial flow compressor at the Naval Postgraduate School at six stator stagger angles between 23.8 and 44.3 for a fixed orientation of the rotor blades. Maximum efficiency and pressure ratio were measured at a stator stagger angle of 31.8 . Results at other blade settings showed that by varying stator stagger angle with flow rate optimum efficien- cies and pressure ratios can be achieved over a wide range of operating conditions. The results of the analysis were compared with the measured results, Suggestions are made for improving the manner of adapting cascade test data to performance predictions. By applying a non-dimensional deflection coefficient it could be shown that minimum work input corresponded to maximum efficiency. The test compressor has a tip diameter of 36 in. and a hub/tip ratio of 0.6. The blading tested is of the free-vortex type with a design degree of reaction of 0.5. Tip speed was about 185 ft/sec. ARY Thesis by .§fuceLC. Marshall entitled: "Effect of Stator Blade Orientati the "Performance of an Axial Flow Compressor" on on ERRATA SHEET Page 17 17 28 28 59 78 100 Line 3 5 6 17/18 20/21 9 Fig. 15 Fig. 37 Change g/cm temper at ve assessories di a- meter analy-ti cally delete "(g/cm3)" Note 2: "DEMENSIONS" Maximum 3- Stage Effiency To dimension! ess temperature accessories diam-eter analyt-i cally DIMENSIONS Maximum 3- Stage Efficiency TABLE OF CONTENTS Section Page 1 Introduction 21 2 0. N. R. 3-Stage Axial Flow Compressor 25 3 Flow Rate Calibration 33 4 Measurement of Compressor Performance 38 5 Performance Prediction Program 43 6 Discussion of Results 50 7 Conclusions and Recommendations 63 8 Illustrations 64 9 References 111 Appendix A Calibration of Instruments 113 Appendix B Flow Rate Calibration Details 122 Appendix C Performance Measurements 135 Appendix D Details of Prediction Program 157 Appendix E Proposed Design of an Improved Inlet Duct 196 LIST OF TABLES Table Page A-l Calibration of BLH Torque Meter 116 A-2 Calibration of 0-12 in. Hg Bourdon Tube 118 A-3 Calibration of Inlet Pitot-Static Tube 120 B-l Precision and Time to Damp for Various Lengths of 128 Capillary Tube in Total and Static Pressure Lines B-2 Listing of Program ONRFLO 129 B-3 Sample Output of Program ONRFLO-Input for Run 14 132 B-4 Sample Output of Program ONRFLO-Ve loci ties for Run 14 133 B-5 Sample Output of Program ONRFLO-Calibration Constant 134 for Run 14 C-l Investigation of Axisymmetry of Flow 143 C-2 Listing of Program ONRETA 144 C-3 Output of Program ONRETA, Run 1, Stator Stagger 146 Angle = 27.8° C-4 Output of Program ONRETA s Run 2, Stator Stagger 147 Angle = 27.8° C-5 Output of Program ONRETA, Run 3, Stator Stagger 148 Angle = 27.8° C-6 Output of Program ONRETA, Run 4, Stator Stagger 149 Angle = 23.8° C-7 Output of Program ONRETA, Run 5, Stator Stagger 150 Angle = 23.8° C-8 Output of Program ONRETA, Run 6, Stator Stagger 151 Angle = 31.8° LIST OF TABLES (Cont.) Table Page C-9 Output of Program ONRETAs Run 7, Stator Stagger 152 Angle = 31.8° C-10 Output of Program ONRETAs Run 8, Stator Stagger 153 Angle = 35.8° C-ll Output of Program ONRETA s Run 9, Stator Stagger 154 Angle = 35.8° C-12 Output of Program ONRETA, Run 10, Stator Stagger 155 Angle - 39.8° C-13 Output of Program ONRETA, Run 11, Stator Stagger 156 Angle = 44. 3 D-l Listing of Program AXC03 160 D-2 Sample Output of Program AXC03 173 D-3 Output of Program AXC03, Case 11, fi . = 0.000 177 Dm in D-4 Output of Program AXC03, Case 12 , c . = 0*006 178 0 Dmm D-5 Output of Program AXC03, Case 13, C^_ . = 0.008 179 Dmin D-6 Output of Program AXC03, Case 21, c = 0.000 180 Dm in D-7 Output of Program AXC03, Case 22, C . - 0.006 181 Dm in D-8 Output of Program AXC03, Case 23, C . = 0.008 182 D-9 Output of Program AXC03, Case 31, C^ . = 0.000 183 Dm in D-10 Output of Program AXC03, Case 32 , C = 0.006 184 Dmin D-ll Output of Program AXC03, Case 33, C . = 0.008 185 Dm in D-12 Output of Program AXC03, Case 41, C^ . = 0.000 186 Dmin D-13 Output of Program AXC03, Case 42 , C . = 0.006 187 Dmin D-14 Output of Program AXC03, Case 43, n . = 0.008 188 Dmin LIST OF TABLES (Cont.) Table Page D-15 Output of Program AXC03, Cases 51, 52, and 53, 189 Dm m D-16 Output of Program AXC03, Cases 61, 62, and 63, 190 All Cn . Dm in D-17 Program CHECK, Listing and Output 191 LIST OF ILLUSTRATIONS Figure Page 1 Compressor Installation 64 2 Accessories to Compressor 65 3 Compressor Casing 66 4 Rotor Blades, Drum and Shaft Mounted in Supporting 67 Strut Assemblies 5 Assembly of Rotor in Lower Casing 68 6 Compressor Sectional View 69 7 Location of Radial Survey Holes 70 8 Detail of Free-Vortex Rotor 71 9 Detail of Free-Vortex Stator 72 10 Free-Vortex Stator Blade 73 11 Rotor Stagger Angle Adjustment 74 12 External Stagger Angle Adjustment 75 13 Inlet Bellmouth 76 14 Inlet Pitot-Static Traverse 77 15 Detail of Inlet Pitot-Static Traverse 78 16 Three-Hole Probe Detail 79 17 Permanent Pitot-Static Tube 80 18 Permanent Pitot-Tube and Three-Hole Probe Ahead 81 of First Rotor 19 Pressure Readout Instrumentation 82 20 Pressure Selection by Scanner Valve 83 21 Simplified Pressure Selection System 84 22 Traverse Carriage and Three-Hole Probe 85 Behind Third Stator LIST OF ILLUSTRATIONS (Cont.) Figure Page 23 Relation of Traverse Carriage to Survey Holes 86 24 Torque Meter During Calibration 87 25 Traverse of Inlet Duct Run D-l 88 26 Traverse of Inlet Duct Run D-7 89 27 Extent of Boundary Layer in Cylindrical Inlet Duct 90 28 Temperature-Entropy Diagram of Compression Process in 91 3-Stage Machine 29 Assumed Performance of Blading 92 30 3-Stage Efficiency Versus Referred Flow Rate Predicted 93 by AXC03, Rotor Stagger Angle = 43.8 , Various Stator Stagger Angles 31 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 94 Referred Flow Rate, Stator Stagger Angle = 23.8 32 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 95 Referred Flow Rate, Stator Stagger Angle = 27.8 33 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 96 Referred Flow Rate, Stator Stagger Angle =31.8 34 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 97 Referred Flow Rate, Stator Stagger Angle = 35.8 35 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 98 Referred Flow Rate, Stator Stagger Angle =39.8 36 3-Stage Pressure Ratio and 3-Stage Efficiency Versus 99 Referred Flow Rate, Stator Stagger Angle = 44.3 37 Maximum 3-Stage Efficiency Versus Referred Flow Rate 100 for Various Stator Stagger Angles 10 LIST OF ILLUSTRATIONS (Cont.) Figure Page 38 Maximum 3-Stage Pressure Ratio Versus Referred Flow 101 Rate for Various Stator Stagger Angles 39 3-Stage Efficiency, Deflection Coefficient and 102 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle = 23.8 40 3-Stage Efficiency, Deflection Coefficient and 103 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle =27.8 41 3-Stage Efficiency, Deflection Coefficient and 104 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle = 31 » 8 42 3-Stage Efficiency, Deflection Coefficient and 105 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle - 35.8° 43 3-Stage Efficiency, Deflection Coefficient and 106 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle = 39.8° 44 3-Stage Efficiency, Deflection Coefficient and 107 Pressure Coefficient Versus Flow Coefficient, Stator Stagger Angle = 44.3 45 Maximum 3-Stage Efficiency and Corresponding 108 Deflection Coefficient and Stator Stagger Angles Versus Flow Coefficient 46 Maximum Pressure Coefficient and Corresponding Stator 109 Stagger Angles Versus Flow Coefficient 11 LIST OF ILLUSTRATIONS (Cont.) Figure Page 47 Off-Design Conditions of Axial Compressor Stage 110 A-l Torque Meter Calibration 117 A-2 Bourdon Tube Calibration 119 A-3 Inlet Traverse Pitot-Static Calibration 121 B-l Variable Length Capillary Damping Device 126 B-2 Variable Length Capillary Damping Device Operation 127 C-l Radial Integration for Total Pressure 141 C-2 Survey of Pressure Reading in Peripheral Direction 142 E-l Proposed Improved Inlet Duct 198 12 LIST OF SYMBOLS Symbol Meaning Fortran Units ALE air flow angle after the inlet ALE deg guide vanes b number of blades in rotor or BLDR stator row BLDS c blade chord length of rotor CHR in. or stator CHS 3 CCC volume flow rate calibration CCC ft /sec per ft/sec constant C profile drag coefficient of - blading C^ . minimum profile drag coefficient CDR Dmin r ° for rotor or stator blade CDS c conversion factor for barometric - psfa/in. of Hg pressure c specific heat at constant CP Btu/lbm, R pressure d mean diameter of blading DD in. DW referred flow rate increment, DW lbmTl R/sec psia wa/tTVp, V to A h height of part of annulus area - in. AA L diffuser length proposed in - in. Appendix E LMAX number of stages for program LMAX AXC03 13 LIST OF SYMBOLS (Cont.) Symbol m N NCASE NRUN NUM Bar r3-s sd sda si Meaning Fortran Units slope factor which modifies K0)10 SLOPM for the camber of the blade section slope factor which modifies 0o SLOPM for the camber of the blade section compressor speed measured by TACH rpm electronic counter run number of programs ONRFLO NCASE and ONRETA number of data points per run NRUN of program ONRETA case number of program AXC03 atmospheric pressure barometric pressure static pressure in inlet duct ratio of total pressure behind third stator to total pressure ahead of first rotor static pressure at permanent PO psfa Pitot-static tube static pressure with reference - counts to atmospheric after the third stator static pressure after the third - psfa stator static pressure with reference PSI counts to atmospheric in inlet duct NUM - PA psfa PBAR in. of Hg PI psfa PRATS 3 _ 14 LIST OF SYMBOLS (Cont.) Symbol so td tda tda tig tiga Wa Nth Jdl Meaning Fortran static pressure with reference PSO to atmospheric at permanent Pitot-static tube total pressure with reference to atmospheric after the third stator total pressure after the third stator average total pressure after the third stator, integrated radially total pressure with reference to PTIG atmospheric between the inlet guide vanes and the first rotor total pressure ahead of the PTIGA first stage or after the inlet guide vanes actual horsepower of compressor theoretical horsepower of compressor velocity head after the third stator velocity head after the third stator Units counts counts psfa psfa counts psf (ft-lbf)/sec (ft-lbf)/sec counts psf 15 LIST OF SYMBOLS (Cont.) Symbol qi qil Mol r RN o RSTAG s SMAX SMI SSTAG Meaning dynamic head in inlet duct, below atmospheric dynamic head in inlet duct dynamic head of permanent Pitot-static tube, below atmospheric dynamic head of permanent Pitot-static tube radius in program ONRFLO referred speed of program AXC03 N/^/t^ outside radius rotor stagger angle measured from the blade chord at the mid-radius to a line parallel to the axis spacing of blades 7fd/b maximum limit ( l-Llo)/€/0 determined from Fig. 29 minimum limit ( C~C,0)/^,q determined from Fig. 29 stator stagger angle measured from the blade chord at the mid-radius to a line parallel to the axis Fortran QI Units Qil 00 001 R RN STAGR SMAX SMI STAGS counts psf counts psf in. rpm/*Y°R in. deg in. deg 16 LIST OF SYMBOLS (Cont.) Symbol sy T Bar rq td *td tiga to av V. i Meaning Fortran specific gravity of mercury temperature at barometer TBAR static temperatue at inlet TO screen reading of torque meter TRAW actual torque TRQ total temperature after the third stator total temperature at the dis- charge of an isentropic com- pression from P . to P tiga td total temperature ahead of the first stage or after the inlet guide vanes total temperature at inlet screen TTO peripheral speed at mean radius UAV axial or through flow velocity in compressor annulus velocity after the third stator difference in inlet velocity at VD 16.0 in. radius and that mea- sured in boundary layer velocity in the inlet duct VI Units g/cm' o_ lb ft-lb °R R R ft/sec ft/sec ft/sec ft/sec ft/sec 17 LIST OF SYMBOLS (Cont*) Symbol lad iav V. oav u c SF* w u iad VA VO Meaning Fortran velocity in inlet duct adjusted VIAD for slight compressor speed variations average of four values of V.. at each radius velocity at the permanent Pitot-static probe average velocity at the permanent VOAV Pitot-static probe peripheral component of velocity VU actual volume flow rate VFC volume flow deficiency due to VFD losses in inlet duct boundary layer volume flow rate neglecting VFT losses in inlet duct boundary layer air velocity realtive to the W rotating blade row peripheral component of relative WU velocity weight flow rate WDOT referred weight flow rate WREM Units ft/sec ft/sec ft/sec ft/sec ft/sec 3 ft /sec ft /sec ft /sec ft/sec ft/sec lbm/sec lbra y R/sec psia 18 LIST OF SYMBOLS (Cont.) Symbol Pi T S Ah IS Meaning inlet angle of absolute flow inlet angle of relative flow ratio of specific heat at con- stant pressure and specific heat at constant volume deviation angle of blade row SD part of the annulus area in the A center of which V. is measured iav isentropic specific enthalpy dif- - ference for frictionless com- Fortran A deg A deg GAM _ Units pression APa.s total pressure difference across DELP the three stages *•' 3-5A actual total pressure difference DELTAP Arvs^ theoretical total pressure dif- ference across the three stages AS entropy difference Al + ^ isentropic temperature difference - € 10 n. 3-S from Tfc • to T \ tiga td actual deflection for blade row DFACT nominal deflection for blade row DFSTAR at incidence for minimum profile drag loss three stage total-to-total ETAST3 efficiency deg ft2 Btu/lbm R counts psf psf Btu/lbm R °R deg deg 19 LIST OF SYMBOLS (Cont.) Symbol L \o rr 0" * ew Y OvV 2 4 lb sec /ft lb sec2/ft4 2 4 lb sec /ft CO Meaning Fortran Unit actual incidence angle of SI deg blade row nominal incidence of blade row STARI deg for minimum profile drag loss geometrical constant 3.14159 air density after the third stator air density at the inlet RHOI air density at the permanent Pitot-static probe blading solidity SO average dimensionless deflection TAUAV coefficient flow coefficient, ratio of PHIVA through flow velocity to peripheral speed average dimensionless pressure PSIAV coefficient angular velocity OMEG rad/sec 20 SECTION 1 INTRODUCTION Axial flow compressors have a wide range of application. They are usually matched to a turbine in a set. Both open-cycle sets, such as aircraft turbo-jet engines, and closed-cycle sets for power generation are in service. In every application compressor efficiency is the critical factor in set performance. Accurate predictions of off-design compressor performance must be made during the design stage to be able to make corrections prior to manufacturing. Theoretical analysis of the flow in an axial compressor involves three-dimensional partial differential equations of a complex nature (1). In these equations it is difficult to account for real gas effects such as boundary layer growth on machine walls and blade surfaces. The resulting wakes behind blade rows create non-uniform conditions; hence, it becomes necessary to base prediction methods on experimental data. These data would be obtained best on an actual compressor. How- ever, most compressors have relatively short blades and small flow annuli. Even if only small pressure probes were inserted between the rows of blades, the flow in these machines would be substantially altered. Test rigs must therefore be used that have large dimensions. A first approximation of the flow in axial turbomachines can be obtained in rectilinear cascade test rigs. The intersections of a co-axial cylinder with the blades of a row produce a series of iden- tical and identically oriented profiles which are unwrapped into a plane to establish the corresponding rectilinear cascade. A finite number of straight blades with profiles similar to, but larger than 21 those obtained, is then arranged in a rectangular duct through which air is blown at the appropriate inflow angle. Flow surveys taken ahead of and behind the blade row give experimental data of air turning angle and total pressure loss for various blade geometries. Such a recti- linear cascade is in use at the Naval Postgraduate School. The results of extensive testing of compressor airfoil shapes at NASA have been summarized by Lieblein (2). The usual problem in compressor design consists in selecting flow areas, blade shapes and blade layout to satisfy design specifications. Additional studies are frequently necessary if the design performance is not reached. An example is the study of Vavra which was made when the so-called Clark CSN-1 compressor of the ML-1 nuclear gas turbine of Aerojet-General Nucleonics failed to perform as required. Vavra analyzed the design changes proposed by the manufacturers and made further recommendations (3). However the ML-1 project was cancelled before the improved compressor could be built so that it was not possible to verify the suggested changes by experiments. The so-called inverse problem consists in predicting the perform- ance of an existing machine with a known blading by means of rectilinear cascade data and other experience factors. Such an approach was carried out by Gibbons and Bartels (4) for the 12-stage Allis-Chalmers axial flow compressor which supplies air to the turbine test facilities at the Naval Postgraduate School. Because of the unorthodox performance of the first stages, and also due to small flow annuli and short blade heights, it was not possible to predict the performance of the machine accurately. However, the analysis gave important indications for design changes to improve the performance of the compressor. 22 This thesis makes another attempt to solve the inverse problem for a 3-stage axial flow compressor built by the California Institute of Technology with the support of the Office of Naval Research. The compressor is sufficiently large so that the insertion of pressure probes causes relatively small flow perturbations. The machine was designed to permit variations in blade shapes, blade angles, tip clearances, and staging. The installation at the California Institute of Technology is described by Bowen, e_t a_l. (5) This report also describes a theory of perfect fluid flow in axial flow turbomachines. The results of this theory are compared with the measured data of the first stage of a so-called "free-vortex" type blading. Part 2 of the report by Bowen, e_t a_l. (6), gave a detailed investigation of multi- stage flow for both "free-vortex" blading and a more complex "solid body rotation" blading. Measurements of blade skin friction losses were greater than expected from cascade tests. The growth of boundary layer along the inner and the outer annulus walls of the flow was less than expected. Alsworth and Iura (7) conducted extensive and detailed measurements of flow patterns in a single stage of "free-vortex" blading. They carried out accurate measurements of the blade skin friction losses and the radial distribution of work input. The final report from the California Institute of Technology was a hot-wire anemometer study of compressor stall by Iura and Rannie (8). They observed that stalled flow regions rotate in the direction of blade rotation without changing shape but with a speed that is not propor- tional to rotor speed. After the Turbo- Propulsion Laboratory of the Naval Postgraduate School was built, the compressor was relocated there. Since then 23 it has been used for laboratory courses to supplement instruction in basic theories of turbomachines. The present study was conducted for this compressor because of the possibility of changing its blade angles. For fixed stator and rotor blade angles a highly peaked curve of efficiency versus flow rate is usually obtained with axial flow compressors. At a slightly changed stator blade angle the peak will be displaced somewhat. A family of efficiency curves for various stator stagger angles is expected to yield a flatter efficiency curve over a wider range of flow rates for fixed rotor blade angles. The wide variation in flow rates required of the engines for the supersonic transport has led to design pro- posals by General Electric for compressors where the stator blade angles can be changed during operation of the jet engine. Hence, the thesis topic is of current interest. For the tests it was necessary to obtain extremely accurate pressure measurements. An extremely sensitive bourdon tube pressure indicator was added to the instrumentation. A permanent Pitot-static tube was installed, and accurate flow rate calibrations were carried out. A parallel theoretical effort produced predictions of com- pressor performance by an existing computer program which was modified and adapted for use on the I. B. M. Model 360 computer system installed at the Naval Postgraduate School. The author wishes to express a debt of gratitude to the faculty of the Department of Aeronautics of the Naval Postgraduate School for his engineering education. Particular thanks are due to Professor M. H. Vavra for his enthusiasm and guidance in showing what can be done with that education. 24 SECTION 2 0. N. R. 3-STAGE AXIAL FLOW COMPRESSOR The compressor installation is shown in Fig. 1. The compressor inlet is in the background of the picture. The exit section is equipped with a throttle valve. A torque meter is installed on the shaft between the drive motor and the compressor. The overall dimen- sions of the equipment are given in Fig. 2. A basic criterion of the design of the compressor was flexibility of operation. Each blade row may be removed. The angle setting of each blade is adjustable in 0.5 increments. In order to minimize flow disturbances by pressure probes the outside diameter of the blad- ing is 36 in. For the same reason,, the compressor has a hub/tip ratio of 0.6, giving an inside diameter of 21.6 in. and a blade height of 7.2 in. To permit simulation of conditions in a multistage unit, the machine has three stages. One row of inlet guide vanes simulates the effects of previous stages. Two rows of exit guide vanes remove the whirl component after the third stage. To insure that general align- ment and tip clearances are maintained the compressor casing is very rigid. The outer casing consists of two cast iron half-cylindrical shells bolted together along the horizontal plane through the axis. The rows of inlet guide vanes, stator blades, and exit guide vanes are held by bolts extending through the casing (Fig. 3). Inside the casing a hollow steel shaft carries three cast iron drums for each of the rows of rotor blades. The shaft is supported by two roller bearings whose outer races are pressed into the supporting strut 25 assemblies (Fig. 4). These assemblies rest in close fitting channels in the casing (Fig. 5). The six assembly struts have symmetrical air- foil sections. Figure 6 shows the nine blade rows after assembly. The casing was designed to give maximum accessibility for mea- suring instruments. Six rectangular instrument ports are located in the upper half at 30 from either side of the vertical (Fig. 3). They are placed in the second and third rotor planes, the first, second and third stator planes, and behind the third stator. The ports hold a special instrument carriage permitting detailed flow surveys in any axial plane. The casing has numerous radial survey holes whose loca- tions are specified in Fig. 7. The blades are made of ALCOA 356 aluminum alloy without heat treat- ment. There are thirty rotor blades and thirty-two stator blades per row. The details of construction of the blades are illustrated in Figs. 8 and 9. The blading is designed to have a degree of reaction of 0.5 at the mean radius. The rotor blades are twisted by 49 from hub to tip, and the stator blades by 13 . At the mean radius, the important blade parameters are: Parameter Rotor Stator Camber angle Design stagger angle Thickness-to-chord ratio Chord length The values of the blade parameters at other radii, and the method used for calculating the thickness distribution, are specified in Ref. 5. 20.32° 29.98° 43.80° 28.80° 0.1 0.1 2.60 in. 2.60 in. 26 The point of maximum thickness was set at 0.35 of the chord length from the leading edge. The thickness was modified over the rear 15 per cent of the section to provide a trailing edge thickness 0.02 in. The thickness distribution is applied about a parabolic mean camber line. Figure 10 is a picture of a "free-vortex" stator blade. Tip clearances of 0.020 in. for stator, and 0.037 in. for rotor, are maintained by the use of shims at the point of attachment. The attaching device permits variation of rotor and stator blade stagger angle. Figure 11 shows how these changes are made. The blade shaft extends through the rotor drum at the hub. An adjusting plate is secured to the blade shaft by a tapered pin. The plate has a series of eleven taps on a circle about the shaft axis which are spaced at intervals of 4.5 . Between the adjusting plate and the rotor drum is a fixed sector which is attached to the drum by a pin which guarantees its alignment. The fixed sector has eleven holes which are spaced at intervals of 4.0 of arc about the blade shaft axis. Blade alignment is maintained by a set screw through both holes in the adjusting plate and the fixed sector. If the center holes of the plate and the sector are lined up, the blade is at the design stagger angle of 43.8 for the "free-vortex" rotor. An angle change of 4.0 can be accomplished by keeping the set screw in the center hole of the adjusting plate and moving the blade so the screw is inserted into the next hole of the fixed sector. Small angle changes of 0.5 are accomplished by moving the blade slightly so that the two holes immediately adjacent to the center holes are lined up since blade angle is changed only by the difference in arc between the two holes. A similar arrangement is used for the stator blade settings (Fig. 10). Figure 12 illustrates stator blade settings which are one degree smaller than the design stagger 27 angle. The two blades on the right-hand side of the figure belong to the first stator row. The set screw is two holes away from the center or reference holes. The blades on the left-hand side of Fig. 12 belong to the row of inlet guide vanes. They are set at an angle which is by 0.5 smaller than the design stagger angle. The assessories to the basic compressor will be described by fol- lowing a path from the inlet to the exit and to the power source. The instrumentation of the compressor will be described in its appropriate place in the same sequence. Details of instrument calibrations are contained in Appendix A. The inlet duct is seen in the background of Fig. 1. It consists of a screen, an entrance bellmouth, and a length of straight pipe. Figure 13 shows the large mesh screen which prevents the ingestion of foreign matter from the compressor bay apron. Mounted on the screen is a mercury thermometer used to determine ambient temperature. It has provisions for psychrometric analysis of the incoming air« Immediately behind the screen is the bellmouth. It changes the dia- meter rather abruptly from almost two diameters at the flare, to 36 in. in the axial distance of about 15 in. A cylindrical inlet duct two diameters long connects the bellmouth to the compressor proper. The inner surface of the duct is enameled to give smooth flow surfaces. The flow rate through the compressor was determined by surveys in the inlet duct. Details of the calibration are given in Appendix B. The survey plane is about midway between the bellmouth and the com- pressor. The surveys were made with a Prandtl-type Pitot tube. In Fig. 14 the probe is installed for taking a vertical traverse. It is possible to disassemble the probe while the compressor is running (Fig. 15). 28 It may then be remounted for a horizontal traverse. Traverses at 30 and 60 from the vertical direction are possible. These traverses can be made by unbolting the entire inlet duct from the compressor and rotating it on its cradle. The design of the probe prevents measure- ments closer than 0.25 in. to the inlet duct wall. A honeycomb straightener is installed behind the inlet Pitot tube to equalize the flow entering the compressor (Fig. 14). Attached to the forward strut assembly is an ogival wooden fairing (Fig. 5) to provide a smooth transition of the flow from the inlet duct to the compressor annulus. Pressure measurements may be taken at any one of the radial survey ports described. Figure 16 shows the probe used for these measurements. It is a United Sensor three-hole probe, model YC-120. A central hole measures stagnation or total pressure. A static port is located on each of two faces of a wedge at an angle of 45 from the total pressure hole. The static pressures are balanced on a water manometer to insure alignment of the probe in flow direction. The probe is mounted on a probe holder which has vernier scales permitting radius adjustments to 0.01 in. and angle adjustments to 0.1 . The accurate determination of flow rate is essential to compressor analysis. For this purpose a modified Prandtl type Pitot-static probe was installed. Figure 17 shows the installation details. The probe is inserted in a radial survey hole located ahead of the inlet guide vane row. It protrudes forward into a channel between two adjacent struts of the supporting assembly which is shown in Fig. 5. The probe is approximately at the center of the channel between struts. The probe is positioned in the radial survey hole by a brass plug, and 29 secured to the casing by a simple locking device. Figure 18 shows the probe and its locking device in the background. In the middle of the figure are seen the three-hole probe and its holder located between the inlet guide vanes and the first rotor. In the foreground, a plug, originally located at the same axial position as the probe, has been removed to show a hole through which the probe can be inserted. Pressures obtained by the probes were measured by a Texas Instru- ments Fused Quartz Precision Pressure Gauge which is shown in Fig. 19. The velocity head in the inlet is about 0»6 in. of water. A low- pressure bourdon tube was obtained for the gauge which measures from 0 to 166 in. of water with the instrument reading from 0 to 200,000 counts. The calibration constant of the low pressure bourdon tube was determined to be 240*423 counts/psf of the pressure gauge (See Appendix A.)« The pressure gauge operates both in manual and servo modes, the latter providing automatic nulling of the unit. Since flexibility of pressure selection was desirable, two Giannini Sp-lOlA pressure scanners were connected to the pressure and reference ports of the bourdon tube. One of the scanners is seen in the foreground of Fig. 19. The arrangement of pressures to the 12- channel switches is shown in Fig. 20. Later in the study a simpler system of five valves and two manifolds was constructed (Fig. 21) to eliminate leakage flows that seem to have occurred in the scanners. At this time the original 3-hole probe was placed permanently behind the third stator (Fig. 22). Another 3-hole probe was placed ahead of the first stage rotor. This arrangement permitted direct measurement of the pressure increase across the three stages. A P» . 30 Figure 22 also shows the instrument traverse carriage mentioned before. It was used for a survey of total pressure in peripheral direction behind the third stator at two different radii (See Appendix C.)« It i-s possible to vary the radial locations of the probe with an accuracy of 0.01 in. The probe may be rotated through 360 about its o axis. The carriage can move the probe 15 in peripheral direction, which covers a whole blade spacing since a stator blade channel covers 11.25 , and a rotor blade channel 12.0 in peripheral direction. The locations of the radial survey holes with reference to the indicated meridional angle of the carriage are shown in Fig. 23. The represented stator profiles are at the design stagger angle at the mean radius. The actual stagger angle measured with respect to a plane through the axis is the complement of the indicated angle. A short cylindrical duct connects the compressor to the exit elbow, which is equipped with sheet metal turning vanes to minimize losses. A transition piece connects the elbow to the throttle valve which is seen in Fig. 1. The valve consists of two rectangular metal doors which move in slides. The position of the doors is controlled by a right- and left-hand threaded lead screw which is rotated by a Graham variable-speed transmission. A revolution counter indicates the approximate valve opening. The compressor is driven by a 50 HP Fairbanks Morse induction motor. It requires a three-phase, sixty-cycle, 440 volt power supply, and can operate at two fixed speeds; namely, at about 900 rpm or 1200 rpm. Attached to the motor shaft is a Baldwin-Lima-Hamilton SR-4 torque meter type A. It is shown in Fig. 24 during calibration with 31 known weights attached to a lever. In the background may be seen a Brown Instruments strain gauge readout which is of the standard Wheatstone bridge type. A conversion constant of 4.1565 ft-lb/lb of torque meter readout was determined (See Appendix A.). Under the protective cover next to the torque meter in Fig. 24 is installed a six lobe flux cutter to measure the motor: speed by means of an electronic counter. 32 SECTION 3 FLOW RATE CALIBRATION The permanent Pitot-static tube was installed to provide rapid and accurate measurements of volume flow rate. At a given throttle valve setting this probe was calibrated against the result of two traverses of the cylindrical inlet duct. Through-flow velocities in the flow annulus ahead of the inlet guide vanes were varied from 110 ft/sec to 70 ft/sec, so that compressibility effects could be ignored. The objective of the calibration was to determine a cali- bration constant which gave the volume flow rate when multiplied by the velocity measured at the permanent probe. An initial survey of the inlet duct was made at the highest pos- 3 sible flow rate of 512 ft /sec. Figure 25 shows a plot of the mea- sured velocities of both horizontal and vertical traverses. The smooth velocity profile in the duct is evident. Figure 26 is a graph of the measured velocities in the inlet duct at the lower flow rate of 3 310 ft /sec, which shows a higher degree of scatter. For these tests the pressure indicator was operated in the servo mode. Even in the meter mode the indicated pressure oscillated considerably. Pressure fluctuations seemed to have been initiated by the vibration of the brass rod on which the traversing Pitot-static tube was mounted in the inlet duct. The pressure indicator corrects a disparity between the angular position of a mirror of the bourdon tube and the indicator dial at the rate of the full scale reading of 200,000 counts in 120 seconds. With the oscillating pressure applied to the sensitive instrument, oscillations of 50 counts were observed for a pressure 33 corresponding to a reading of 600 counts. To reduce these oscillations, variable length capillary tube damping devices were installed both in the total and static pressure lines of the traverse Pitot-static tube. Details of the arrangement and the results of tests are shown in Appendix B-l. A special survey was made to determine the thickness of the bound- ary layer on the walls of the cylindrical inlet duct. However, con- struction of the traverse probe prevented taking readings closer than 0.25 in. from the wall. Figure 27 is a plot of the readings at high and low flow rates. In each case it is evident that large losses occur near the walls. With the velocity distributions of Fig. 27 the flow rate is about 2.5 per cent smaller than the value obtained without considering the changes near the walls . Data were taken in the boundary layer at intervals of 0.25 in. from 17.75 in. to 16.00 in. In the main stream, data were taken at 4.0 in. intervals from 16.00 in. to the centerline. With horizontal and vertical traverses, four points were obtained at each radius and two at the axis for a total of 46 locations. At each station the following measurements were taken: q. - velocity head in the inlet duct (counts) q - velocity head of the permanent Pitot-static probe (counts) P . - static pressure with reference to atmosphere in inlet si r duct (counts) P - static pressure with reference to atmosphere at permanent Pitot-static probe (counts) T - temperature at the inlet screen ( F) P - barometric pressure (in. of Hg) Ba r T - temperature of column of mercury ( F) Ba r 34 The barometric pressure P (in. Hg) was converted into pounds Bar per square foot to obtain the atmospheric pressure P by A PA - P8ar (71. 4G7) (Psfa) (1) where the constant was obtained by correcting the height of the column of mercury for temperature variations in specific gravity of the liquid. The inlet temperature was converted to absolute temperature, and was considered to be the total temperature T because of the low veloc- r to ities, or T,D = \ - 4^7 (°R) (2) The static pressure P at the inlet was converted from the meter reading to an absolute pressure by A £4-0.42.3 where the constant was obtained by the calibration procedure described in Appendix A. The pressure P measured at the permanent probe was obtained in a similar manner. The local air density p. at the inlet was computed from _ O.OOZ^aPc W = lL^xl0-) (lb sec2/ft4) (4) and similarly p at the permanent probe. The velocity head in the inlet q.. was converted from the meter reading, q = ^ (psf) (5) b>1 240. 413 The dynamic head at the permanent probe q was obtained similarly. The velocity V. in the inlet is then J i Vt -V ^ o%°X (ft/sec) (6) 35 A similar procedure gave the velocity V at the permanent probe. The arithmetic average of the 46 values of V obtained during the carrying out of the traverses gives the average value V of ° oav Voa^ = J£ Z-i V°j (ft/sec) The measured values of V. were adjusted to account for variations in flow rate due to slight compressor speed variations Visd " Vt — (ft/sec) The four values of V. , obtained at each radius were averaged by 4- V.av = "5" Z- V^j (ft/sec) J-1 The volume flow rate was first calculated by neglecting the losses in the boundary layer to give N^ j. . In general the volume flow rate is obtained by V r Jr (ft3/sec) If the above equation were integrated directly, the velocity at the axis would not be included in the volume flow rate calculation because the zero radius is a member of the product under the integral. Therefore the integral was replaced by a summation. \Ft = H V.'av *A (ft3/sec) The quantity A A = 2 Tf r h/144 represented that part of the annulus area in the center of which V, was measured, where h is the height iav of the area. This method assumed that the velocity measured at the radius of 16.0 in. exists also at the wall. A volume flow rate 36 deficiency was then calculated to account for the difference in velocities at 16.0 in. radius and those measured in the boundary layer, or VD * Vtav ,fc. - VCaV 9>L- (ft/sec) By a trapezoidal integration the volume flow deficiency vV becomes , Jfc.O \F0 s I 7f I VD r Jr (ft3/sec) 7 /fc.O Then the actual volume flow rate is ^"c - ^"t ^p (ft3/sec) A calibration constant CCC for each run was obtained from ^C 3 CCC = — - — — (ft /sec per ft/sec) oav A first series of calibration runs was invalidated when a leakage hole was discovered in the traversing Pitot-static tube. After repairs, the calibration constants CCC were obtained with a maximum relative error of 0.1 per cent. The final calibration constant was 4.4050 3 ft /sec per ft/sec velocity obtained from the readings of the perma- nently installed Pitot-static probe. A data reduction computer program, ONRFLO, for these calibration procedures^ was generated for the I. B. M. Model 360 computer. A listing of the program is included as Appendix B-2. Sample output data from the final run is given in Appendix B-3. 37 SECTION 4 MEASUREMENT OF COMPRESSOR PERFORMANCE The objective of the study was to determine the effect of changes of the stator blade orientation on compressor performance. It was decided to restrict performance investigation to the three stages only; namely, from a station ahead of the first rotor to a station after the last stator, or between locations SP-2 and SP-8 of Fig.. 7* The effects of the inlet, the inlet guide vanes, the exit guide vanes, and the diffuser were not considered. For tests at a particular stator stagger angle the barometric pressure P was determined first. For various settings of the Bar throttle valve the following measurements were taken: T - ambient temperature ( F) o P - static pressure with reference to atmosphere at the permanent Pitot-static probe (counts) q - velocity head at the permanent Pitot-static probe (counts) P . - total pressure with reference to atmospheric between the inlet guide vanes and the first rotor (counts) A P - total pressure difference across the three stages (counts) T - reading of torque meter (lb) N - compressor speed measured by electronic counter (rpm) With the relations listed on p. 35 these measuring data establish the following quantities: P* = ?b&r C-7/.4C7; (psfa) (1) Tt. ' T0 * 4 5^.7 (°r) (2) P - p^° Ko 240 4 11 <3> 38 = 0.00^78 PQ 5,8.7 . _?^ ^U^) (lb Sec2/ft4) (4) (»Utt)(l44) T\0 Ti0 bo1 240.4£3 V. -J ^^ (ft/sec) (6) The absolute total pressure ahead of the first stage, or after the inlet guide vane, P . , is tiga' p. . p. - Ln "9« A i^Til (ps£) (7) The pressure difference across the three stages A P is obtained from aP »•«* = 210. -Ui (ft-ibf) where the constant was obtained by the calibration procedure described in Appendix A. The angular velocity CO is with the measured speed, N ZTf N CO b — (rad/sec) 4,0 Both volume and weight flow rates were calculated. The volume flow rate v ^ is then \fc - V0 (4.4050) (ft3/sec) (8) where the constant was obtained by the calibration procedure described in the preceding section and Appendix B. The weight flow rate w is 39 . .p 32.H4 P* 518.7 .c- P» ;., „„,, w = c o*wxi«)Tlo = ^ 17, U 88fe) rr 3„s - — (ii) and the 3-stage total-to-total efficiency n ,_<. obtained from P rWfh 13-5 = r (12) r WA where P..,, and P„ are the theoretical and actual horsepower, respec- Wth Wa r > r tively. The theoretical power is P , ,given by PWfh = w A^isn78.3) (ft-lbf/sec) where Ah. is the isentropic specific enthalpy difference for a frictionless compression from P^, to P , after the third stator, r tiga td Assuming a perfect gas there is Ak;s - cp A 1*3.3 (Btu/lbm, °R) where AT ' is the isentropic temperature difference from the total temperature T . ahead of the first stage, to T . after the third tiga ° td stator (See Fig. 28.)- Since T . very nearly equals the ambient tiga temperature T , there is to 40 K = cp Tto / r ' ta With the isentropic relation 1 (Btu/lbm, R) r, td -td T p. the final expression for the determination of the theoretical power is = W C. 'W+h vv "\> The actual power is obtained from \0 [( Pr3.s) y- IJ77S.3 (ft-lbf/sec) (13) »WA = Ipq U) (ft-lbf/sec) The total- to- total efficiency at the three stages is then a 3-S \v cp lt< [ttvS)?- l] 77 8.3 (14) For comparison with other compressor data certain non-dimensional parameters are introduced. The average flow function Toy *s fc^e ratio of through flow velocity V and peripheral speed at the mean diameter U , or with av V* = 451.4/144 = "e l*«0 (n/sec) and, for the mean diameter of 14.4 in. U. ftV CO 14,4 (ft/sec) there is AV V^c (i2)(K4) CO (l4.4-)C4S|.4) NTc (JO (0.184Z) (15) 41 An average dimensionless stage pressure coefficient 7 Av/ is defined by the program prints out "surge in blade row xx" and proceeds with calculations at an increased flow rate. The positive value of ( C~ ^,0)/£ ._ at which this occurs is the input quantity SMAX. If C^/C^_. exceeds a value of 2.0 at r n J D Dmin incidence angles L smaller than C ( 0 , the program prints out "minimum pressure in blade row xx" and halts computation. The negative value of 45 (C-L|0)/&10 at which this occurs is the input quantity SMI. In both instances the symbol xx refers to the number of the blade row, and indications are given also whether the row is a stator or a rotor. For any incidence angle within the useful operating range, the values of the deflection ratio and of the ratio C^/C^ . are computed by sub- D Dmin J routine CASCAD and control returned to the blade row subroutines ROTOR or STATOR. The efficiency of a blade row is determined by subroutine ETACAL. Losses due to tip clearances, secondary flows, and wall friction are taken into account with the relations proposed by Vavra (1), If conditions of surge exist, the flow rate is increased by about 1.0 per cent and the entire computation is repeated. If the incidence angles are in the so-called useful operating region of Fig. 29, the output of subroutine ROTOR or STATOR becomes the input for the sub- routine that calculates the next row of blades. Interstage data are printed out as the program progresses from one stage to the next. If all blade rows of a machine have been processed, the overall compressor efficiency and pressure ratio are computed for the particular flow rate. This flow rate is then increased by a specified amount, and the calcu- lating process is repeated. If the so-called minimum pressure is reached, the computation stops and a summary of the overall performance parameters is printed out. The details of the calculation of the flow through the inlet duct, the inlet guide vanes, the exit guide vanes, the diffuser and the dis- charge passages are not described in this thesis since it is concerned only with the performance prediction of the three stages of the com- pressor from a station ahead of the first rotor to the discharge at 46 the third stator. The methods applied for the flow analysis in these passages may be found in Vavra (3) and Gibbons and Bartels (4). According to Howell there occur additional effects that influence the off-design calculations. The growth of boundary layers on the walls of the annular flow channel will change the velocity profiles and can reduce significantly the effective flow area. The program takes account of this effect by a blockage factor that can vary from stage to stage. This boundary layer growth is responsible also for increasing peaks in the radial distribution of the axial velocity components in succesive stages. These higher velocities outside of the wall boundary layers produce smaller incidence angles for the main portion of the flow, thereby decreasing the actual flow deflection and reducing the work absorbed by the fluid in the stage. Hence in a compressor that consists of stages with identical bladings, the last stages will produce smaller pressure ratios than the stages at the compressor inlet. Experience shows the efficiency in successive stages is not reduced by the peaking of the velocity profiles, and their effects are usually taken into account by a so-called work-done factor that is about unity for the first stage and gradually decreases for the successive stages. Such work-done factors can be introduced in program AXC03, but for the present compressor they will be taken as unity because of the small number of stages and the large blade heights. For the same reasons the blockage factor is assumed to be equal to unity also. The geometry of the blades could be obtained from the information of Ref. 5. Values at the mean radius were used for the analysis. Bowen (5) determined that the average flow angle at the mean radius 47 after the inlet guide vanes ALE was 20.0 for the design inlet guide vane stagger angle of 11.4 , independent of flow rate. From Fig. 12 it can be recognized that the inlet guide vanes were set at a stagger o o angle of 10.9 for the tests. Therefore a value of 19.5 was used in the analysis program for the average flow angle leaving the inlet guide vanes. Reference 5 does not give values of C__. for the Dm in blading. It is possible to introduce in the program various values of C__. both for rotor and stator blading. Hence it was possible to Dmin ° r estimate an average apparent value of C_ . for blading by comparing Dm in measured results to predictions of the analysis program for several assumed values of Cn . . Dm in A prediction of the 3-stage efficiency for different stator stagger angles is shown in Fig. 30 for an assumed value of C_ . of Dm in 0.020. The light line connecting the individual peak efficiencies represents the calculated operating envelope which could be obtained if continuous control of stator stagger angle for maximum efficiency were possible. For comparison with measured performance a series of calculations were made for stator stagger angles of 23.8 , 27.8 , 31.8 , 35.8 , 39.8 , and 44-3 . For each stagger angle, values of C^ . of 0.000, Dm in 0.006, and 0.008 were assumed. To avoid confusion with the computa- tions of actual performance based on measured quantities, which have been called runs, each calculation of the analytical prediction program has been assigned a two-digit case number. The first digit identifies the stator stagger angle of the case as follows: 48 First Digit Stator Stagger Angle 1 23.8° 2 27.8° 3 31.8° 4 35.8° 5 39 . 8° 6 44.3° The second digit identifies the assumed value of C^ . of the case Dmin as follows: , ^. . Assumed Value of C^ . Second Digit Dmin 1 0.000 2 0.006 3 0.008 For example, case 42 is the predicted performance at a stator stagger angle of 35.8 with a value of (L . of 0.006. The summary output of Dmin each case is presented in Appendix D. 49 SECTION 6 DISCUSSION OF RESULTS Experimental data for performance measurements were taken during 55 hours of running time. Another 81 hours of operation were used for miscellaneous calibration and testing, including 44 hours for flow rate calibrations. The rotor stagger angle was set at 43.8 for all runs. The design stator stagger angle is 288 . For the performance tests, the stator stagger angle was set at 23.8°, 27.8°, 31.8°, 35.8°, 39.8°, and 44.3 . Measured performance is presented in comparison with the predicted results of AXC03,and also by establishing dimensionless performance parameters. Figures 31 through 36 are graphs of measured performance in comparison with the prediction by AXC03 for each stator blade angle setting. Figures 37 and 38 are summary plots of maximum efficiency and pressure ratio for all stator blade angle settings. Figures 39 through 44 are graphs of the non-dimensional parameters at each stator stagger angle. Figure 45 is a summary plot of maximum efficiencies and associated deflection coefficients for all stator blade angle settings. Figure 46 is a summary plot of maximum pressure coefficients for all stator blade angle settings. On each of Fig. 31 through Fig. 36 the 3-stage efficiency reaches a peak value for each stator stagger angle. The referred flow rate at which this peak occurs decreases as the stator stagger angle increases, Also plotted on each of Fig. 31 through Fig. 34 are the results predicted by AXC03 for the three estimated values of C^ . of 0.000, r J Dm in 0.006 and 0.008. The predicted 3-stage efficiency reaches a peak for 50 each C^ . • The referred flow rate associated with predicted peak Drain efficiency decreases with increase in stator stagger angle. The referred flow rate at which the predicted peak efficiency occurs decreases with increase in Cn . at a given stator stagger angle. AXC03 predicts the condition of surge for each case according to the criteria of Howell shown in Fig. 29. The surge condition is assumed to exist whenever the parameter (C~6.0)/€.q exceeds the maximum positive value SMAX which has been inserted in the program. The incidence angle for the rotor blade rows becomes larger as the through-flow velocity decreases until a surge condition is indicated. AXC03 predicts surge at the same referred flow rate for each setting of stator stagger angle. This is due to the fixed angle of the inlet guide vanes. The assumption has been made that the air flow angle leaving a blade row does not change with variations in flow rate in the incompressible flow regime. This result was verified by Bowen (5). The inlet guide vanes were not rotated as the stator stagger angle was varied. At all stator stagger angle settings the first stage rotor had the same incidence angle at particular flow rates. In the stage-by-stage analysis, "surge in rotor 1" was indicated at the same flow rate in each run, regardless of the stator stagger angle setting of subsequent stators. The region of referred flow rates smaller than that at which surge is indicated by the program is marked with a dashed line on Figs. 31 through 34, and is less than 41.5 lbm y R/sec psia for every case. When the value of the parameter (£""*• /p)/^© ^s determined to be less than SMI, the negative value of the parameter which has been inserted in AXC03, the program stops the computation and prints 51 "minimum pressure in blade row xx." The referred flow rate at which computation is stopped decreases as stator stagger angle is increased. At stator stagger angles of 31.8 and lower (Figs. 31 through 33) the referred flow rate at which computation is stopped is larger than the maximum flow rate of the graph. At a stator stagger angle of 35.8 (Fig. 34), the computation stops at a referred flow rate of 45.5 lbm "y R/sec psia. At stator stagger angles of 39.8 and 44.3 (Figs. 35 and 36), the program did not establish useful data. At the high stator angles the referred flow rate at which computation was stopped was lower than the referred flow rate for surge in the first stage rotor. The latter flow rate remained unchanged at a fixed inlet guide vane angle. This difficulty illustrates some of the limitations of program AXC03 when applied to an actual machine. The program assumes that the compressor is designed properly. Changes from stage to stage must be gradual. For example the annulus area may change gradually from stage to stage with no adverse effects. The checks for minimum pressure and surge limits are applied in each blade row. Surge in any blade row is interpreted as surge in the machine, and similar limitations are imposed for minimum pressure. The program cannot cope with discontinuities similar to those that occur by leaving the angle of the inlet guide vanes unchanged. It is possible to imagine a situation where the angle setting of the inlet guide vanes might have to remain unchanged; for instance, if the inlet guide vanes must support the front bearing of the rotor shaft. In such a design the first rotor would have higher blade efficiencies than succeeding blade rows at large flow rates^while at lower flow rates 52 the blade efficiencies of the first rotor would be smaller than for succeeding blade rows. The 3-stage efficiency of an actual machine is an overall parameter which includes the effect of the different blading efficiencies for different blade rows. In an actual machine the flow will adjust itself to these conditions, whereas program AXC03 cannot cope with these peculiar circumstances. For these reasons the program did not produce performance data for the higher stator stagger angles of 39.8 and 44.3 . The detection of the actual surge point during the tests is difficult with the available instrumentation and must be based on acoustic phenomena. Incipient surge, probably due to rotating stall* was associated with an unmistakable oscillating change in sound level produced by the compressor. This effect appeared suddenly even when flow rate was decreased slowly. To eliminate the sound due to this surge condition it was necessary to increase the flow rate by about 5 to 10 per cent. A flow rate slightly above the surge point could then be reached by gradual throttling. For each run one test point was taken in the surge region and one test point as close as possible to the surge point. Efficiency at test points in the surge region was below 0.90 for all stator stagger angles. The efficiency at the data point close to stall rose from 0.89 to 0.94 as the stator stagger angle was increased from 23.8 to 31.8 , and efficiency decreased to 0.89 with further increase in stator stagger angle to 44.3 . As the surge condition is approached the rate of decrease of efficiency is set by the relative location on the efficiency versus referred flow rate graph of the two data points in the surge region and close to surge. At a stator stagger angle of 31.8 which produced the maximum 53 3-stage efficiency and at 35.8 , efficiency decreased most rapidly as the surge condition was approached. For other stator stagger angles the efficiency decreased less rapidly. It is not possible to make a meaningful comparison between measured surge point data and the surge condition calculated by AXC03 because surge was predicted to occur at the same referred flow rate for each case. It is possible to compare measured values of maximum 3-stage efficiency with the maxima calculated by program AXC03. For all four stagger angles at which computations were made, the maximum 3-stage efficiencies were predicted to occur at flow rates about 5 per cent greater than the flow rates at which the measured maxima occurred. According to Howell the point of maximum efficiency occurs for a value of the parameter (t~^,0 )/£/o about equal to +0.19 (Fig. 29). At greater positive values of the parameter the losses are believed to increase rapidly. Since the measured maximum efficiency occurred at a flow rate which was about 5 per cent less than was predicted, an analysis of off -design performance of a com- pressor stage was made. The method of Vavra (Ref. 1) assumes that the air flow angles leaving a blade row do not change with flow rate in incompressible flow. The velocity triangle changes caused by a reduction from design flow rate are shown in Fig. 47. The subscript d refers to the design condition, and the primed quantities refer to off-design. The analysis was conducted at the maximum efficiency point for a stator stagger angle of 27.8 . The following quantities were measured; Design flow coefficient Cp^ = 0.517 Inlet flow angle o( (fixed) = 20.0° 54 The off-design flow rate was fixed at 5 per cent less than the design flow rate, to correspond to the flow rate at which the measured maximum efficiencies occurred. The off-design flow coefficient is 4> = (0.95) (0.517) The peripheral components of velocity are: ■^ D.?5" tan «, The relative inlet flow angles for both conditions are: v ' 0.15 Av , the efficiencies follow the same pattern as when plotted against referred flow rate, since Eqs. 10 and 13 differ only by a constant. Figure 45 is a summary plot of the maximum 3-stage efficiciency for each stator 59 stagger angle versus flow coefficient and is similar to Fig. 37 for the same reason. Figure 46 is a summary plot of maximum pressure rise coefficient for each stator stagger angle. The pressure rise reaches a maximum at a blade angle of 33 and a flow coefficient of 0.475. The curve of corresponding stator stagger angles has an almost constant slope. Theoretically the deflection coefficient Tw represents the change in peripheral flow components in a blade row divided by the peripheral speed. Figure 47 is a stage velocity triangle showing design and off-design conditions in a compressor stage. At the design condition d = U = U An assumption in this analysis is that the air leaving angle from a blade row does not vary with incidence. Hence the angles o< j and /3 remain constant. The relation between the off-design deflection coefficient L and the off-design flow coefficient Y is fixed by similar triangles 1-6 ' " C. + d and t = / - ( i- rj <*>' or ^. is expected to vary linearly with

«^ ±_-f i^_— * fc- t + 4-J 4 + + fl =31— t £ •4- -4- 4- +" + + + z < <0 It! K) o 66 FIGURE 5 ASSEMBLY OF ROTOR IN LOW^K CASiNG 3 K UJ > -J < Z O UJ CO S CO CO Ul o CO UJ CO 69 ^7 CO UJ _» o UJ > D CO 5 3 o § o J ui QC © u. 70 •I 5 Si UJ UJ UJ QC CD 71 o s a: o Id LU K U. U. O UJ O UJ 72 OJ a < oo a: o i- CO X LU cc o > I OJ LU UJ u. 73 z UJ s (0 -> O < UJ z < UJ O o < K (0 (E g O O 74 FIGURE 12 EXTERNAL STAGGER ANGLE ADJUSTMENT .0 if •■;;: ::::"i|^n FIGURE 13 INLET BELLMOUTH 76 FIGURE 14 INLET PITOT- STATIC TRAVERSE 77 JL H I-MH u (5 I UJ u. O _J UJ Q in UJ IT D © «*J *fc» g 78 FIGURE 16 THREE-HOLE PROBE DETAIL COMPRESSOR CASING BRASS PLUG K"7 9.60 T I 3.60 3.35 4375R- i__i -2.0875 3.0625 0609R FIG. 17 PERMANENT PITOT- STATIC TUBE 80 FIGURE 18 PERMANENT P1T0T - TUBE AND THREE HOLE PROBE AHEAD OF FIRST ROTOR 31 FIGURE 19 PRESSURE READOUT INSTRUMENTAT ION in en UJ > < UJ 3 Q. TV LU Z UJ OD < 2 or UJ O 1- Q_ Q. @ ®' O j < f- i- < o h- H- (/) ▼▼ UJ OD O or a. UJ _i o X UJ UJ cr x © @e" (in) cm io _> y S£ < i- t- F < < O *- *~ t. (/) (/) ® Ul or < UJ > or -l or Ul < < o: O i- CO ▼ ▼▼ T ▼ GO O LU _l UJ - t CO 0> Q »- s 3 "5 !2 .5"° 2 °- r^ki o -o Q. 3 z o |St££w L^vJ • 0, C ££C0!-O 3 2 •2 <- o «/» 3 LU * £ _| Q_ CD LU CO o LU r^ti cr CO CO F £ U^vj 0) 0_ LU ^_ .O cr C 3 Q_ • a Q !° LU o> ._ 0.0. O 3 Q- h£i ll Q_ L^sJ in Q. w z O a> CO Q. ja A 3 c\i • 0) t? S 2 S2 0) 0) Q_ r^ti 2 3 s £ QC CD u. 3 TJ > w |P c o> 22^ »2 O w « c 3 1-Q.m _ o 84 FIGURE 22 TRAVERSE CARRIAGE AND THREE — HOLE PROBE BEHIND THIRD STATOR 85 Si * * $ CO i Id > fit D CO < 1 UJ I o UJ Ul i 86 FIGURE 24 TORQUE METER DURING CALIBRATION FIGURE 25 TRAVERSE OF INLET DUCT RUN D-l VOLUME FLOW RATE 512 FT3 /SEC 75 o 50 (0 u o • 25 > ^>0-0-CM Left 15 10 Horizontal Traverse 5 0 5 Radius, in. 10 15 Right 75 50 o CD £ 25 u o Top Vertical Traverse 15 10 5 0 5 Radius, in. 10 15 ^ Bottom 88 FIGURE 26 TRAVERSE OF INLET DUCT RUN 0-7 VOLUME FLOW RATE 310 FT3 /SEC. 5 60 40 o 9 2 20 Lift Horizontal Traverse 15 10 5 0 5 Radius, in. 10 15 Right o o > 60 40 20 0. Top 15 • 10 Vertica I Traverse 5 0 5 Radius , in. So 10 15 Bottom 89 c 3 s tr £ 1000 800 600 400 - 200 - Udocity ft/tec 60.06 — 48.04 36.03 - 24.02 12.01 18.0 17.5 17.0 16.5 160 0 16 0 I6.« 17.0 17.5 18.0 Radius , inches 90 A«, BTU/lbm-°R FIGURE 28 TEMPERATURE-ENTROPY DIAGRAM OF COMPRESSION PROCESS IN 3- STAGE MACHINE 91 o Q O V a o o CVJ ^--^^ t S i , CM 1 "e t cd fc 1 1 f \ £f \ \ .6 ° 5 -J O X i i Xx CD Z 5 < 03 / V CD \j >\ * ,-020 FIGURE 3f 3-STAGE PRESSURE RATIO 8 3-STAGE EFFICIENCY VERSUS REFERRED FLOW RATE STATOR STAGGER ANGLE =23.8* PREDICTION BY AXC03 FOR VARIOUS CDmin Maximo CDmjn =-0-0.000 -P-0.006 -A-0.008 Test Data — o- W(lbm/sec)-(T^CpR)/PA(psia) 40 Referred Flow Rate K rs 94 - 0.80 0.75 35 1. 015 FIGURE 32 3-STAGE PRESSURE RATIO 8 3-STAGE EFFICIENCY VERSUS REFERRED FLOW RATE STATOR STAGGER ANGLE =27.8° PREDICTION BY AXC0 3 FOR VARIOUS CDmjn Maximo CDmin =-oo.OOO -D-0.006 A0.008 Test Data — O— W ( Ibm/tec ) 'TlV> (J*S >/R(ptia) 40 . JTZ Referred Flow Rate W to 95 0.80 0.73 35 FIGURE 33 3 -STAGE PRESSURE RATIO 8 3-STAGE EFFICIENCY VERSUS REFERRED FLOW RATE STATOR STAGGER ANGLE* 31.8° PREDICTION BY AXC03 FOR VARIOUS CDmin Maxima CDmin =0-0.000 -0-0.006 0.008 Tttt Data- W (Ibm/Mc) «|t^ fcPwiJfptlo) 40 Rtftrrtd Flow Ratt W 45 96 -0.80 0.75 1.020 - - 1.015 - FIGURE 34 3-STAGE PRESSURE RATIO 8 3-STAGE EFFICIENCY VERSUS REFERRED FLOW RATE STATOR STAGGER ANGLE = 35.8° PREDICTION BY AXC0 3 FOR VARIOUS Comin Maxima Co^ =-0-0000 -q-0.006 -6-0.008 Ttst Data— o- W(lbm/Mc)jT^ (>f7R)/PA (ptia) 35 40 Referred Flow Rate W {ho *A 45 97 3-Stago Efficltncy 0.95 0.80 3-Stogt Prttturt Ratio 1.025 1.020 — 1.015 — FIGURE 35 3-STAGE PRESSURE RATIO A 3-STAGE EFFICIENCY VERSUS REFERRED FLOW RATE STATOR STAGQER ANGLE* 39.8* Tost Data— o— W (lbm/toc)0 (J^/P^psio) Te»t Data— O- 1.015 0.75 15 40 Referred Flow Rats 46 W W 99 0.80 0.75 35 FIGURE 37 MAXMIUM 3- STAGE EFFICIENCY VERSUS REFERRED FLOW RATE FOR VARIOUS STATOR 8TA00ER ANGLES PREDICTION BY AXCO 3 FOR C0rnn * 0.006 T«st Data — O- Pr« diction W(lbm/i«c)4Tfo" U*R)/PA (psia) 40 J% R«f»rr«d Flow Raft WV -j i Sl LOO FIGURE 38 MAXIMUM 3-STA6E PRESSURE RATIO VERSUS REFERRED FLOW RATE FOR VARIOUS STATOR STAQ6ER ANGLES W(lbm/8tc)fl^ (^VR. (ptia) 20* 30 REFERRED FLOW RATE W-k^- 40 1 101 3-Stogt Efficiency, ^3-S 0.85 0.25 FIGURE B9 3-STAGE EFFICIENCY ^ ■ 3~S DEFLECTION COEFFICIENT *bv , AND PRESSURE COEFFICIENT y,QV .VERSUS FLOW COEFFICIENT ov STATOR STAGGER ANGLE =238° 0.20 °n 3-S ov 0.40 0.45 ov 0.50 0.15 0.55 0.60 d> . Flow Coefficient 102 3- Stage Efficiency, ^3-S 0.95 0.90 0.85 0.80 0.75 0^9—1 *o © Tav Toy 0.35 0.30 — 0.25 FIT3URE 40 3 - STAGE EFFICIENCY 7J^S DEFLECTION COEFFICE NT T^AND PRESSURE COEFFICIENT \^ov .VERSUS FLOW COEFFICIENT <£av STATOR STAGGER ANGLE = 27.8° 0.20 V T( 3-S av ^V< av 0.40 0.45 050 0.55 0.60 ^av » F,ow Coefficient 0.15 103 3 -Stage Efficiency, 0.95 0.90 0.85 --0.80 0.75 FIGURE 41 3-STA6E EFFICIENCY T) 3_s, DEFLECTION COEFFICIENT Tw , AND PRESSURE COEFFICIENT \f/w VERSUS FLOW COEFFICIENT^ STATOR STAGGER ANGLE? 31 .8° 0 ^3- 3-S 7 T( "av fa, 0.35 0.30 0.25 0.20 _ av av 0.15 0.40 0.45 0.50 0.55 0.60 , Flow Cotfficitnt 'OV 104 3-Stagt Efficiency, ^3-S 0.95 FIGURE 42 3- STAGE EFFICIENCY TJ DEFLECTION COEFFICIENT TL AND PRESSURE CO EFFICIENT v/fc¥ VERSUS FLOW COEFFICIENT ^>av STATOR STAGGER ANGLE =35.8° OV3-S AVI/ Tfl av 0.40 0.45 0.50 ~1 0.55 <^v , Flow Coefficient 0.20 — 0.15 0.60 105 3 - Stagt Efficiency 0.95 0.90 0.85 0.80 FIGURE 43 3 -STAGE EFFICIENCY ^-s DEFLECTION COEFFICIENTT^AND Tav PRESSURE COEFFICIENT^ VERSUS \f/ FLOW COEFFICIENT d>nu STATOR STAGGER ANGLE = 39.8° - u.»0 B,?3-S VTav Af0 0.30 — 0.20 0.20 0.15 0.40 0.45 0.50 0.55 (k , Flow Cotfflcitnt r» 54 Run B-l 6>^ w*- Calibration Constant 240.141 Counts Per LB/FT2 0 20 40 60 80 100 120 140 160 180 Instrument Reading, Counts x I0~3 o ^ • s 12 14 16 18 Calibration Constant 240.612 Counts Per LB/FT2 6 8 10 12 14 16 18 I Average Calibration Constant 240.423 119 TABLE A- 3 CALIBRATION OF INLET PITOT- STATIC TUBE RUN C-3 TOTAL PRESSURE MINUS STATIC PRESSURE PROBE 2486.25 2157.5 1413.75 1002.5 793.75 TUNNEL 2508.75 2182.5 1421.25 1011.25 797.5 RUN C-4 STATIC PRESSURE BELOW ATMOSPHERIC PROBE 2450.0 2190.0 1422.5 1087.5 795.0 TUNNEL 2455.0 2190.0 1425.0 1012.5 810.0 120 FIGURE A-3 INLET TRAVERSE PITOT-STATIC CALIBRATION 3000 Probe Error q Probe -q Tunnel q Tunnel -.01 0 1000 2000 3000 Calibration Tunnel Velocity Head,Counte 3000 o a. a *» CO o ct c 3 o o 2 a o S 2000 10O0 Probe Error Pq — Pq- °Probe °Tunnel pk w •-Q < in » V. — » (NJ o fM IT* *w C < o in a> *^» (VI c fM >— r- •O -0 o— n mo *-» LU >-> z k- > » <•* 3>a: 00 — Z r«- «/> ••< 3 "■* • — » x-— ■•-^ uoo no nD •Ot- 0.C? p— i — a __v » •- » LP. 1— 1 v. O— ■ x M.O. * a. —H LTiQ< ZZ CO ^s a ■— >cC >— «-i h- — 1 o 3 ►a. 31— ' m — ■«-. O— . - C?oO C\i o h- »ou ► CLO O W QC ^ -~r— too — -. »c a h- 3 t— 1 '_>—< f*? Z — O • ji 00 **+ ;~. ~oo fM •— za — Ql 5; — >.£ »• ►-4W ^ N» CO # — i «.w .— I 00-4^ — . — i/; — — II a.om s0 * a ioa MZ »- *> • m -~ h- ••in 'j1' 00 » * • CD 1— — -— c <~ zza! ITi f— >— jjc ii O*- w,-< . t— i ro ►— ro h- .n »>a z— — x ~a >t r*- — i om-uma — z a: O — k-s)- • CM ^sOH-^OO H-iLTMJJ — —•.—.—. o ■w—. -H CT «■ ->^r^~ \oo 3w »a>0- i-»xT v0~fNJ +• •»■ iT> • -C7-7Z -~ O J_C ^ nun-— I •> nJ-J-c\jC7^>d- — ZoO-— Z* J" Z>ONNNO>t —. ►ONOU — f\J w-«w -K- a ■— * ^ ^ 3 «-N- ■» •> -h-r\i ii^r- »co# + — ,T|-HoO>-< -H ~_< 2Z <<— II II CJ 11 II ii — <«— ii -^wf~ •- — »-xxxa:xJO II <— 3Lbo:— •Q.|7»Q —j i— i r-i fM ro i_i , 1 >■ J.J -J O > •— « X 1 j: — < :NJ "O >j-ir\or*-aocr0'-*('\i'0->>> + + ♦ -f 30C30 <<<< fM "V+ + + + > > < o > > •«• — 31- + > a ii > ii ^ uo ~- • « t • lf\ rO >*■ LA vO _J st-vt'vl'^rOi— ir*— *-* > t "-"-f •• •— <^-^ ►— « *— * *— t — I^v«- •»■•»* > .oo • •+ + + + -f- •■r\| ♦ •f\l»o 3 •■«■•»••»•■«■ fM —* O 7" on X X-tt- ir.LTiLPilTi— ~»— • — — — > — — — (— i— WOOQQO z— . -»~vo^a^ i->t vt>j-<<<<< — sD^-fO^vj-Lri^-^-inf^— <.- 4»^>— ii-ii-i_ii-ict> ■J" ►— LT\ CM IfMTi »lT-H • • • »>>>>>>t r* || -r-< (. »fSJ »^4J- ■»■ ^ » ► II -»0 II >D-0 II O • • • • • II II II II II -£ Z<;-Z ^~.m(Nir\)(\jAj — — UJ LUUJ 'X! II II II II II <— — •— • — -~OJ LT\ O I— vC h- 1— I*- (— —.-.—.—.—. i-t f\J rT) nJ- LO h- <■_, »— i. — < i-.^-4f\jrr(Nj"iri— — — <-"—■«-« Onfioidoa: — <<<<< cc s fa O O •H •P in •H 0J PQ CD H ■9 EH in v0 _! o j'j'JOOO'JO^^fH^^^^^^^fNjrrj^fvirM^rNifNjrxi^rorof'iro^fO^^nfl^^^^.f^^^ J oOOOCJOC.JOOOOOOOOOOr^OOOOOOOOOOCJOOOOOOOOOOOOOOOOO 3 -jaocjo^ooooooooooooooooooooooooooooooooooooooo 129 sC ro > QC_JZ^- < ^ ac—XuJ3z ce. -O QS->03 X if\ H-v>— OO —4 V * >*■>*■ >*->*- >*•>*■ LSION LABO CALIBRATI R.P.M.//5 ,7Xf 14HPR0 5X,12HHEA0 HPRESSURE, X CO a z a u UJ CO 00 OLUOo »v0 f\J o^-4<\irnsrmso a>-o»-co^ a , — v r\j -J- >t >t vt >t >t >t 0<>>>>>> COO N«|/) Oi— IZK H- LL' LU oj DU-XUJ «»z LL t- +++++++ h- mjiiiD < a:-zzo » O t— •> O C> 00 r-- vD m >*• ror\j(\i(\j(\j(M(M QOQOCQO <<<<<<< z o U- E SCHOOL 0MPRESS0 NI3,27X, ,7X, 12HI CEiMTERLI HHEAD, C CO LU t— o a -j o O — . • > i— o3>Xrg LU z + 4- + + + + + r\j • « < c£*-X>*-* > «H ■— — ICN -» Z)-J —-^ ► O j_ V..-I z 0 r»-cC0NO'-< • «1»^««*W— —>— ' --in > ••» 0< Oj*voO < Z z a QOQOOQO i-KaQ.oiJiujJJ30Jiji:ui2:(j UJ —I O > < a: ci-j-ir'vor^coo^O^fNj rn^j-LnvOf^-cca^c-^fMmvj- m\Or,»coo>vO'^'\im>j-LO-or^acc7-0'-^rsjr^>}-mvO^-ooov t— acccx D*C0CC u^O^CT OsCTCTCrc^ON(j^OOOOOOOOOO'^^^^^^^^^^(\l<\irsjr\jf\|(\ir\jf\)r\jf\i or cooooooooo ooooooo^-H-<-^^'-<-^r-<^^-j---^— i-<-<— i-j^-<^^-<^^_j^^^^-i o oooooooooo ooooooooooooooooooooooooooooooooooooo U. 162 M o o o LU o ^^^^•^^^%t,^^^^^^^»rir.iniririi'sirifiiPifNin«co»e«o^f*-^r*-^r"-^^^"^-co rooccoccccococcccocooocooooooocooc oooooooocoooo OOOOOOOOOCOOCCCOCOCOCCCOOOOOOOOOOO OOOOCOCOOOOOC cooooocc ooocoooococoooooocoooooooo oocoocooooocc COOCOC C-NCOCOOOCOOOOOOOOOCCOCOCOOO oooooooocoooo O OO- >tiriririvCC COCOOhOCOOOOCOCCOOOOCCONCOOOCOOOOOCOO •v CD o V. o o CM rvj — «f> I- II < < c coluxz zlui«* •-» ac~o UjOmJJZ dIIC0j< r-m < OC ►•— 'QCH-LU Ok »-oi/i_j 00 »^0 Zv. Z iu>frinoc <*v oc a. »v. z — oco Q-f-^ joo-woi- O II 03U.I-1- »CI- o c~oi-llo • cocoo OLL OLULL'OZ -JLU II < JZ zz -z uj t— -lu>zok:3:d O LU OC O OOCOOt/Ot// LUQC JZUH «/» t- OC >< LU O H> Z Z LU LU ~-ac _iO>— •— acoc f^_JLL«— XLULULU Q.O. • !3LUO_IUjLU J— X ll z c>— ••— • Q-KU <•- ocll LL LU ujc CM X o CM z QC O romi*-* «~K"I •C oz < » T. X LU CL LU QC >«: cDK-i_LUCL 11 ho3 ii Lb< II II — r-tr* M — II 11 5- mc-i LULLLU f-i— QC OOO «~-J<»- LU <<< vt-JQC< a. •»!—»— I— • < OC —•^000000 mac* oc -w -*LUOLL\ O >vO LL>_JOC CO — OOQC ►DU.3 lA *vQCO<<< in i/> uj voi-i-Ki- • xco oc — (flhO<«t fO OOLU CL "V l-U- LUOOO •4-3L5 r^LOLULUZl— h-l— • oo<> _j a.~ < vs • < o QC --i & <-> LO O ^sv^00_J_JO »«h •• • •— i Q r^.rOLLOoCcO 3tX *X LO LL LU • • >*■_! (<1X«1 — i ll. >— r — r- »o •»<*"! •• U-O LU —• <_>N*-r^r*-u. oooo »-o j^-.^K.rr, QfQCQCoCOQC < ••w.v. <<r rOZCCiOQC(M'Nt ► •■r3QCOCQC_ILU_l lAQCoOULU • 0Ax_j**z>caocr>o'"-f\tLr\ | ►— cm h-rocx LU ►— r-i<- LL. •• •■QCH-QCO*— ► » •> • m«0 LULU • » — .-tX>tUJUJUJOOXXXX>- 100 LUU-'»XQC-c\jvr oooXfNJfM • •■ »LU »«*v OO •— tLUXXXCM vJ-lo «r- » ••O'— ••-'QOLUPnenrrfn mv. r«- ••o^t>>0'>Cf\j X - «-r» LL f»- •-^'ZcC LU • • • • • * LL — I »••>••»»»•> >H* Is-! s: xsiS'-^Cj r-r-u-oLux lu .ocdcocccdxooxo xox •• »2: 5T Z 2T 5T 51 2f + + ►-.►-ON^r^.-i^CLLUQ.oO LULULULUvOLUfMZ>->0— •^|-f\Jf\J(Njrgrsjf\|f\J(Mf\J^-i«-< c — r-ts-~-i-i-i-i-oz>-i-i-cji-i-i-(-»-t-i-i-i->-h- □ luluwCJw— .-q_ll, n ii i-uysriKi! liirxsrsxrj-zrrkxmiijrrrjTJ «— ^-Ol— l-UJLU<v>LnO2;OLULULULULULULU LU LU LU LU Z LL LU LU LU LU LU LU LU LU LU LU — if\lfO>f irivO -^r\i T-* _j o i-H (NJ r<^ ro xj- lf\ vO f^ CD CT O •-< fM-J"0 <-• t\J rO >}■ in vO r^ CO cr •-• r-h-r^r^f^-r*- r»r^r-r"~r-f^-h- r^-r-r^- r — r — r — f — r~r— r— r^-r-r~r— o v m o o o £ o bD s •H •P CO •H H ■3 En > z j-Lnsor^coo^O'-<<\ir^>rijn^ois- O OOOOOOOOOOOOOOOOOO LL cocro-^fMr^vt >J- >!• IfMfMTl LOLO Lo-'PU<'0»a"LnvOf>~cD ifiu". ^C ^0 O-O "0 O O sO vO ooooooo ooo ooooooooooo 163 c o o LL O LL —> «vj rvj r\j (\j f\u\j v _/0> < co < » O am Of »v LU • LL' O >- vO_J<— CC _J LU LLi-xa:>-o O o o CO CO o p o o a1 ■H 0<0)l-Z •■ +£ 3 XO LL"— < vC OUJU — •"—iluotoqc > f\J-I_JLU — <>OLUO 13— •— ' z r»-o_i-i>L5 »o »- , — LLk-X' •• < X < (/> OoO •• J-LUQC 2 * >-< £ LL -I — XO OCO — H rvjxro • Q. t— ••< i •■(Njinc-'C xi— O •■ -ZCCCC Z300 _j >(MX LUX — ZCX 2 — LL O •<_ZQ.CNJ^-"-' *£ — •— -O^O NV-o" Z — «- — — LL »q:lu ►fMLOvO'^ ►>*U>»0 O>0^ ^J3!_i-r-^ ii Oi-KN o O nCnC— *00qC\JLL •■•••■ I *-_J— V^LU — ..>0-0>CX LLC^OO — I"1"! a I — ►— t— — >■»— — X— II—— z C -- >— o » cooQ^ojaaazc'— volujcilz o ll ll u. »ll oi:i"iz:ox>x:Ej-Lr\ o o -^ o-^ LU -0-00 vO o — < oo > r-r^-r- r- o o oo uj cm mm o > z ~-*~-t r* ■ ' . — t . — * .^ — ' . — I . ( - — « . — < 1 — < ~~* — A CJ ooo o ooooooooo ooo 164 c c < a. ifi «o «o h- oo o- o <-* »o r- oo a- o ph o r- oo & o <-< OOCOOOOCOO«"*<-"«-"i«-i«-'<-'<-i"-«<-i'-*f\l ► • V »-O0 CC oo m< •> O -JO o ^ «z _j o - •■ < o -f\. 5" •4/) — O0 lug. a. •> Q •■ <-J •— • .— • a LL < f\l »- •• 00 ro * t a 00 *■ LL o » o fO vD O0(\J 00C j£< Qt O0|- CC O * O LL -¥. <■ -[±1 ~ oo ^ «. » aoo x *-c£> ~Z - ~ •■ r\j occ i- • cr a 0'-J"m>ot*-oc <3> o^f\j H- O OOOOOOO 0»-i-*-*«-«<— '-*i-i--i^-*r\|(\jr\J(\j(\|(\|(\JC\jr\J f\j f^rOf<*l 0£ O OOOOOOO oooooooooooooooooooo o ooo a O OOOOOOO oooooooooooooooooooo o ooo 165 o o p ii »-> »a! » » ,» <; 00 » << X O -*- o z • s |_ h-< # •■ LU~H O -^ LL .h ZLU — • O (\i fV « O-^ O OOO -~ •■ <•«• Pvl LTi < « _J O — iQ- »- •■•• -joo i-»oo^->r^ "y I- CCX O KO LU LU •«• — O. • » LL Q o »uj a; ooct cc co x^h#o c\j» oco»- — * H <0 CO00LU«~ 0 "§ 3:0 fM«- * * fcOOin t- OC OO CO O fSJ LL H «—i_) oor-m— o^x>— s» # in 00 c\j o: •• z,«v_j--iooiriooatLJvt x— co qc «\i —or, m ^ ►- ♦ c— 1 ucrin^-xi— «-ir. ir.rsm o-^~* 3>— «— o _i ►— CvQ. >—»-'^:NsaN»-<>- 00^:0 a^ 1 cr oluinjcm «. oj >v m »- •> CC - 1— i>t<>— ifM •»— — > »->J" T — -~CC ^ o^-<>-Z3 •m>-ox •■ »^ • _j— eo 1 •»■ • at » >rx luoo z ••» ji^» .cr<om ltv — <>rx;x C •— •■ L_5-JKZ*_JLUO0LL-T ^Z# tO0 ^ O »-LL ffl ^^ QXHBKQ- 0«3 ii ii ii — »_j>x* ii ii ii lo ii >t-i«:H-h-osr zj 03-- <^-^OII_J_l< II (NJINiO II _l II II CD f\Jf\JX0^0SQU.mi:ilL «-«<\JO -- r-i — * -J_J o-* o^>*m LUo >r-t vt^>r>}- o c o o UJ o «r rsj m •$■ it o r- cd oo & o «-« j Pvc co rvi r\j m m m m n*> <*". m m m m «rf- »fr «$• -4" »$■ •. •■ » • V fs. OC oc o cc uj •* * •— ' O CC — . — - —I ►UJ X i— ■ i-l UJOO —to QC u. C< oc •> < f\i •o << »— » iz >-a: oo m ^ O •> OOOO •> • . ►*: U, •> K CO -P X - OO •> U. c >*• *-m »oo o » 9 r\i -t~ oc - — o en Js (\i XCo © f^o oo »-m oo u. < eg y n ooo. -a: - » •>. 9 » ► — CC< UJ » u, C_J» ••► C oOf\j rem 0< — •■ <• ► •>< LL- CC O — OO k- m OO < •■ rn X uj • JE»-QC r-< _J •■ *vO ^ -(^ ^-CD Z » • CI N ri •■ "-oo cc:» . -h mi-"— m ,3 *r "-qc »oo i UJ # — > O mrv C -JTh- ooo ^» # # CI in < i • h- QCOO< » » (NJro # -» 1/1 O > N ""i < •• ■+- l-O _J_I ~ Ci H ► • U. f*l K- O_J00 00Q£ <3 < <\J00*O PO» oo o ► -~ -<. Z «- •> oot— i^tr>oot\i a) ►ZK • QOQ.» H i£ccOOLU«~ <_J ^PO_j«.»o rg — •■ «oomLL h i3C^oor-(M— > cmmo h- •. i— x o cr ec < r- .-t »< > < » •. m m>}-xi— »ir. ir.jrm crv — ♦ t- •«• <— o _j — • ►— ••k-»-i^'vcr'- i cr ujf^uj a; v in ~- •» ooco«— ■Nj-o fv »r\i — -f + «a o — . — .oo •■O— <►— ^3 »PO>-OX •> «-^ • _l~< I • • QC * .J-X UJQZ ••«■ >r~~c£«a'0'-< — < •» o o-o-C >vZ-M- < O (_)■»••»•+•«• >J-O0 vO O ••LLf'l I— f\J QXMBKD- 00o^oo •■«■ ro ♦ — «■►— 2: c_ o o. uj •» rxi< x >-o<< it ii ii-j_j>x-& n n n m ii >t_^h-(_ ox rs aj 3oo_j ii !D_j5.cj<— 33><>Qn.l- CLQC>XOiCOjlOU.m^ctUJ -J_J O — < Om^J-LTN uj< oo mooo >0 >OnO >o^ooo o j- m vt) r~- co a 0'-^(\irr-^in'Or--aDo>0'-jr\|(M(Mf\j(Mf\J(M f\J ropnro 2C O OOOOOOO OOOOOOOOOOOOOOOOOOOO o ooo O O OOOOOOO OOOOOOOOOOOOOOOOOOOO o ooo a. 166 c- c c u < a c m 00 o o o (V • or c m^»0 fVJ<\jf\J DOC coo ooo ooo ooo oooo oooo oooo oooo oooo o o o o z ** CO 3 00 UJ o — < to o LL CO o < •• o CO * o UJ CO CO LU>~ or> O- c • < U-X Oo CO c CO X uO >• a: c UJ X X> o co X3 CO CO Z-O cm >- I 1-0- >— CO o 2< cu o< 2 CO O0 O I 2.2 LU ZI 2 i—UJ CO LULL t-O «I _JQ no ox _lt- o 2 u CO 5" -< (NJO ^•m •<* o>e I • coo * I co< * * o> o «-" <\j m «* in >o r- oo a* o «-" <\i ct «$• >Of*-Kh-r^r^r^r-r>-r^r^oDcooooooc <\J(\J<\Jf\J(\lf\J(\Jf\]f\l<\)<\l(\J<\JM<\lf\l oooooooooooooooo oooooooooooooooo oooooooooooooooo oooooooooooooooo oooooooooooooooo f\i i r-o • ^ * *<\JO <\j o oo »lu >fr — I I 0* LUO m i sf LU o^m OCsJ •r»- 0(M + • coo * + .-co o* I CO LU* — 1< vO* ^o LU o LU o u 2 cc o t- o < LL 2 D u LU or a o o CO CO LU X o c> o^ o • CO trv cr ft c cc + fVI ro*. o* • X o»- II * x-^ l-O — m LL^J- CO • om i—ao LLsO < I a. o 2 < O co CO > 2 LU a o-*co>t vOrc-M-O^ •-ir^.cO^ r-in» • tfMT.<0 • •* I OCt# CCINjCMvC NO^O -• I <-< I • LU »LU oooonO I O + co cocsjcoa* ♦ cc*^ cr •* • hotO-j-o^ w*o i r- LLINJ | LO-* oolu* r^ O l mooro "-■Lurg* • ll co co coo LL < X a o z < o o CO LO > > LU G U<~iU(JUU^ ^->UU U^JU CD < o I o a LU a*t *»ooo ~*r-tr>j I 00 »LU I I opj LULU + ^0 r- co co co irr»-*cr> cn«i-cof\j co-*» • lTl»H<0 • .# + 00*00 ♦ + ♦# oooofot t<-^ in •* • ooO>0# * 00 I ^IA< r-ujvoo* -Hh->0 I < OvO •LU* • i-^OCX oir> + so* 1 vO0 <* o + >r •—^-4 I 001P LLOLU*r*- o I r-coi*- olu^* • "-t^a-coo LLC0-J-* I (M < CC o or o x u I CO CO LU Z o co > 2 a > LU O or o o z o o LU CC CC o o CO 00 LU Z o r^ >r m >c r^ co o" o ~* og i*^ >4- «t «t CCMJ(MJ,a,aOOOOOOO (VJNM(VJIMWMf,(OC,Cl^mCl oooooooooooooo oooooo oooooooo oooooooooooooo oocooooooooooo OCOOOOOOOOOO«-«r\j (T\ I COPsJ OUJ*0» I mcoa cum* • COh-» I IfMfX-CO 0 •* * r^oto r^o »lu co I oao couj-f in ro m co in • o*<* o*lu** < 0(\JO*<00 o«~ • OtNJ** *x —• | ros0< int— n lu— + eo* »-o »-^t00 CO II CnJ oo<^ i n ^r ii O I * 'LU »-XCL >—o r^- oocro o o oo^ o o ooo o o ooo 167 o c c IX' ir>or,-cooxosoNo-oso~*fvJrri>frLo, (f, fTl (*> fCl f^ fO fO fO CC pn fC (T| OOOOCOOOOCOOOCCOO ooooocooocccooooc coooooooooccooooo coooooocooccooocc *v X c o o or o IL o >* <\J <\J (\J o f- • >c o II z 1—1 LU X 1- t- < t— . o 2 •z O »-o o Of •» * c LU5» f\J 5" ir oca c a. X (NjU."— U-xon oot-o •• ii lu_j< — Xll.—«oll.k-_jOoou-ujz c> oi/ioo<0 3:llj o fVJ rr\ O O rvj f\i > ►— ^-i,— i^h «— . — i •— ' -^ t— < .-< <\j *• • m * « X « • o o >i- KM < r^ Of 1 O 1 o sC r\i x mi mm LL O0<£>00 I 4 rvj oo oo *r^*LL m r» » o 00 00 »O0m (\J CO vD £» c> o 1 X c LL •— * ^<00<— * mm II 00 LO K ©**rvj • » 00 *— ♦ 00 < LU | # LL * • C\J •» o 1 r- a: oooo<»« X rvjm ,3 oo 0. tm I LU -^#* <* * ♦ m O LL LUOO o z eo<<0 o xir\ < O u ta* LU f^# » vO Lf\ * • • H 3 o •» z< -J o F-iMHrooo r- ir— . oo o MM o— 00 M OfMOO* o O o m X m# H oo LL > 2_ + MH< Q. m m •» ♦ >»■# «■• ocz o Lurom^ oo ««» c » » vO o >-*-ooro<» o -J »-4fri m o m o«- < CLLU «• >- <\j^00# 00 oo CO oo o m fVJ •r* u _J D —I (T • **c* ^ * mm m » m o>r 00 LUCD a a »-•+ 1 ooo oo «• o •• •■ m irt • ♦ o < OLU V. Q. 00<«l 1 # < mm 000DO«Ni m o o sUCT o <•« »— * • »»LU< — ■ oo> LL 'OO o m 1 ON>r U_l (_> oo II <\J»- • Z 0 ( 0**- LLLU + OLLOOOmX X • »m »_JOLLO + O ooo ooo L- — »j-lu »ir II CDQ + Q 1 II 00 Zoo zoo OLU* O •>* z o o< LUO o z LL0C>J- | O O | O0X— I CC 1 ac i ii o ii o> i ^* Of 2 aro LI < o xor^oo i o-< ii uu.oJ"-<-)»LU O II > II o — 2:— <-»m»-— r\it-— _j -J^sJLL || II X--L-0 • 3 ► _JLU »- fc- *acoo<-4 mm* • LULL II II LLLLLL l| LULL II LULLLUOLU II LL < # II LUZ o ooz OX LU o LL 1 -<•«• (NJ OOQC mm u a. < O «-TM rn.fr ■— i _) -ILL o u 0fr COLO >or«- UJ LU UH 00 mm mm mm mm -J uuuuuuuu o z -cooN0'-'i-LC>cf~ coocooooococ ooooooooooco oocooooooooc oooooooooooo C OOOOOOOOOOOO "V GO o — V. I o * O -J I o < • "B sj- t— -1 II u o X LU -1 0* < (M *^ * l/> ♦ o • o OLL. • >* CO o o < m. ... — CO # t/1 S) to ojq: • Ph >C f\J * cm _< — ~ ►00 > LU C aJ + O '_rf _JCC — * -J ► •— o _i —ex — ■» < _JC + LC l_> <►- t\J — *-*♦ Eh <0 LL1 -JLU • <_>z — <— o>-< >j-fc-'Nk LULJJ< <#C> — LUU. — Z — 00 + ^'V 21LZ< — l~LU O — LU<»— kNULU I I— *-*LUa£Osv< • O Z>o0# LU<0» QC + — Z O CLiJ«~> I IS) »#<—**&. 5 ■*.*-> II •— MO< II 3 CD< II < II II 0-»ljlu>-ujci[:i-u u _J -J LU < > o > z 0'-|cf*-oooso^<\jr^>'friri»or-c>ON^r~h-r~h-^h-r"-^cococooocoeoaocooooo©vfj> cooooooooooocooocooooooo OPOCOOOOOOOOCOOOOCOCOOOC ooooooooooooooocoocoooco ccoooooooooooooooocoocoo a oooooooooccooococoocccoo ■v 00 o o o UJZ >o UJK >t l-U <\! Q.»- • CM -0C — » r- UJU. <\i vO 00 ax X UJ II •+- UJ*- 1 * * UJ O^T < »- h- oc < ►3 a. a oc -o— i < — ► nC Ui N ►QC I at sr>o O 00 | 3Q£ * x^o — H- • — < — < »— - •• m i — • xuje; i— ■ O -* ► vO Xi— — ujo;«-' 22 •«■ *t— o -ac LUUJ • • • >- w — imO xo a:oi-a. Z.CC-+Z t— 1 —>sx ox »o a «oa.wQ.w o — Q. — UJ + 1 o t— <■ "C I o 1 + o* * *v t- WZZW cc- - 1 <- «i— o-:5: • UJllJ*- • o rvar (\ja: ho:cfiz o OUJ<(\J — 3:—' 00QXCLC laoi-oaaa X «:i-o ii II H II li — < n i— a. n >-rMa H h- ii oo n n 3 m< it — i r\jrr •—< II CM ~< II — < LU II UJUJI— O -p O o OO 8 i o o •H +3 w •H 1-3 H ■s i_j2xxxiaiiou-aOiu.aLiflajwi-iiJ2 ooOUJUjujujQ.CL^ao,-'Q-co^ — a.ot— >Q-Q-acuj o _j o — '<\J rosj-in ■£> < o oo ooo o O —I •— 1»— I Hr- I.-I t— I <1 sc — i f\j m >}• m *o r»- oo o O >— ir\jro>j-in\Cr*-ocO" o<— rvjm h- O COOOOOOOH"*'- !-*■— i ^h _i_i_i .— ir\j (\| CM (\J cc o oooooooooooooooooooooo o o oooooooooooooooooooooo 171 o o o o O'ff'fO'^iMro'ooccoococortHHHHi- cooooooococooccoooccoooo COCOCOOCOCCCCCCOOOOCCOPO coooooocococ occocoooccco cooooooocooccocooooooooo o rocccoocooococcooccocooo 00 o "^ o c UJ -"■ U0 — t- o, »Z X ULiC UJ ^-^ ►->- # Q-t- # -P MJ < § .* OJ— a: 5 rv' >ar Q. vJi O, ►IL I r- uj 1-4 oo o p o IP w 3 Ji oot xr Q-t- UJ ii «■■— * <] lu » v < ~ llO t- ►oo ~ +i oou. — # ~ aa. < — • frM « _ _l MQ - I ? C oo — oo ^ 25- "7 l_J »-Z— Q- I U-< Q cc •••— ill «— s. ro it. ^ O a. — i ^ < o o — — a> »— UJ • O O to LO »-~*v 2 t— 00—< < — •■ ► — ' LUX aj ••H-w- a. "vX-O O I oo< eh JCQC» * I5^U O Zh-O ~- s- -~ — O — - - ~-»-» oor.o zoo *«J-fv voio «->—•«• QCJ I LUh- »00 <_'0 •*>—-— a.oo • >^— fMLoio— mo •»• *~\±t O0-H — » «- .— o — C •— <*OOLU ijjLU— »H w-*Ojo OJO OOOCO*— 00 Zo£^Z>-va O — ■Q.LL'UJ + O' »— LOQ.+0 Q-|0-»» •— •» h- ^^ZOO Qt— I CC *~CC 0C-aC..Z o nx'aaauj o _j _i o — 'oj ro^io o jj < o oo ooo o > l_> lf\ LOlO LOlOlO LO > < c< _j r\i m >t io >o r>- oo O^O-^ojr^vj- 10 0^-000^0-^01^ »• o oooocooooooooooooooooo a o oooooooooooooooooooooo LL 172 IT* • O I o o o o UJ u> a • z o* <-* —i X a£ O o • u. • f- OCO -* UJZ O a. -• z DO Oct CO LOU. mo ae a.H- o u< o UQ lO -it- o < J xz o <.** o o a • 3 P» o m O o a ae o o o o o o • o o o o «* • t\J o • m t-t t • — * o r- — < o o u o o o • o o o o ** • CM o • \e\ -rf • • r— 1 •o r- "■* o o o o o CO m • o o CO o ac ao —> t rg cc o u • ID -* t a • H» o >o CO < UJ CO X ■*• u < -J < u. o O ae o T ir\ >o r«- rn O o <_> o o O a o Q o a o Q o a o Qt cc o ae • ae fM ae o ae o 0 o r» o — < u —4 173 o o m o p fa o £ Ch O o o o o o o fM (M CM 0 49 W5 • • • • • • rg rg rg o CO o CO o • «a • ft < )— «5 h> — o < o ^ o < o < o a o < o a o < o a o Q CO CO <-J GO CO a CO CO • ac • • ac • • ac • ac rg a r» ac rg o r- oe rg o r- o •* h- (M O •T H- f\j o «»• ^» rg t— <1 i— ac rg m • o CO CO • 0* flO o IT O CJ — i rg o — i rg rg O m rg O a o a O u o a o a O a O ce. o ac O ac o ac o ac o ac O <» t ■ « • « • z -I UJO < O — IN «M IM lf\ «*■ in X ar uj •••>•• a. uj_i ^ co ^ eo -• co ,-. o o> «* CD r* -*■ in *• r G* in o rr, o -t o — » o O o o o o o O o o o IM M (M CM r IM in f\l — * r» in 0> r(\ O —4 o fM CO t • • • • • 4" —4 fl — t <*■ -J in <» m ~s> m 4- -r in r- r- r~ r- o o o o O o a- o> C7> o> a* 0" > a UiZ ouj <— 00 »-o m 00 — o 1 u. o I*ML • UJ o < LUCC I o <9 te -cc o> £ O03 Ml 1 00 o rooo • o UJ mi ae ■p a 1 >- s ou zz •r a; — JJ 0> H Q — CO & «o CO 3 UJUJ ro O > u.i_> ^-.-j^^^i o z < ujz i i i i i i o y < a:« X ot: a o a-- u. ai- ot u< uj ec o> oe o < ouz o o>- o— uo «»■ •< ox uo . o uj — ec uj uj a. ho >- _j x «s 30 u < t- oz o x i/> < X C^ _( < UJ t-UJ < O I- LU_I — z _io X - zz < —< a ujo uj q— •- •••••• J ODUJ o— 'O—O— 0~0-~0«~ -• — 3 O V/>«/>l/>V»«/»«/»tVJI/> o >r aco vzcT'ztrziT'zcrzc^z z z _i r- oo momomomomoino o o <. 0K-v0h-'Ot->0*->0H- ►- »- — — ««-««™« in QQQQQOQQ II o oiMZcozr-zr-zuizmz Z z t-O UOOOO0»OOQC»OOO O O (_> ^_j • tij to «o »o "O to o o uj uuj •fir >r >r •j- >r >r uj -> ►->— h-»— N» I— I— t— Q. iJJLUUJUJUJUJUJUJ i-o i— o f^_j-< joo j>r_ioo«jm_i _i _j UJC >T»-Oh-fMH-OH-fNJH-Ot- ^ ►— Q _l_l »3 «3 «3 «J «3 «3 3 o uj zuj joomo at _i_i_/_i_j_io_j uj -jo r~co«ati«-<o O'OOOOOO'J-^HtJ^IMtHMMM _IOJ C7>0000O00000OOO00O zai ••• i-4 o-'-'-«-t-*-i-«-*-'-«-*~it^-«'-«-< H- • _ aia. oooooooooo- U. z 0 ujo o JEi— r- on o O.U. • UJ o o o > o JZ oo ■* < — m QtO ^-4 Uj« CO >LL • on. o UJ r» ec c\ JO r- !/>•-• -— 00 H- o ■JJ< • ccac ■— a UJ UK 1114 o acac o ec o UJZ t U.O •o UJO •f 175 ii o > o UJ > UJ o I 00 < a. • > 1- — r- O < a (Nl itZ a UJ OUJ a. II _i ►-U -j UJ UJ 00 — < O _l 1 u. at ■a o fa. UJ ec z UJ > UJ < □ > < IX o •• UJ — UJ o ►- U 13 < Z < UJCt ■a ►- o C 00 - in -c in 0"ooooooooooxc*oveooop-r»omg)-p^pg'~" COOxO>l>O>OxOxO>O>ON00aO00CD0OO0GDGOCOaOCOCOGD ooooooooooooooooooooooo gjtnp-in.j-gjpggjcogf-co'— gr»^oopn>tr^g^cop»o•$• it -a- (\j o p- f\ o m o -o o m o ur> c* •* oo 0>,cooop-p-g3g3m«»,>»-p>,>mpgpg— — © PSI fSj (NJ (\J — 1-* — —I — < _,__,_. — _. — _l_i_ _1 — — — ooooooooooooooooooooooo -r (\j r- — o» ■* — < m o> p- in m oo in o m G* pg ** >o vj o -t — i h- ao ao co a* oo oo oo oo oo oop- p- p- gj in in gfpi rg —> oo> n> 000000000000000000000000000000000000000000^^ ooooooooooooooooooooooo o pg m ao —• o ■O O* (*1 g3 f"- pt p- p» •*■ oo r>- -o p- cm ao o ■r o gj in pg gj — i pg rr\ ■& pg — o^ r>- in pg ao in — < p- pg p- ci — — ^i— — ooC*oor-vOmi'r>fNJ"'*ooor-o»O^0gj'O'O»O>0tnminminmm ooooooooooooooooooooooo ooooooooooooooooooooooo g*.jf\iP»m<*ir-r'i-ro"><\i--<»>OOOOcor»o -« rg rg -^ »« —i o o o* oo oo r— -c m >r fn •-• tr> r- -o >r -^ a> aocoooaoaoaoaoaoPfcr-rfcr*r*p»r*f^r*-gjsO"0,Og3m ooooooooooooooooooooooo -*OfMO*'-Nroov>fri-omrMaoo-iaDaom"©oaoo<»'i f* ~0 >r oo * ao rg r- o m o» (\i r» O in ao <\i ga co ao oo ao r- r~ gj -O m m ■»■ >T rr\ ro rg -< — « o O 0" co ao r- OOOOOOOOOOOOOOOOOOOOOOO • uj< OOOOOOOOOOOOOOOOOOOOOOO u aa ooooooooooooouuoooooooo ^ oc nonomomomoinomomomoinomom Z U.O ro J-~rmingj JOr-r-COCOO^O^OO-^-^rglMrrirri^-* 3 uj_i •rgj* -r >r gr!■ gr Jt — < P* OU. -J >r ii an. 3 • UJ o Og UJ u _» m _l o «i ii <_) o 2 u> «i UJ >- UJ K o a UJ ->z 00 O -J J- o <— a «I ao UJ ►- UJ — ■x 00 >u. or OIL UJ QC UJ LL u LU >— X U a UJ at JU oO— _u< C£Ct a. 176 IT O 9 -J 1- — m O < a (M LUZ a LL CLU 0. II «a — _i f-O _i a. LU co — <3 O _l 1 u. ar <3 (J5 (ML U-' or z LU > LU «. c > < or O « a1 •—• LU IS 1- O o — (f>CO ijj o < LL a ■* >- o£ • 00 Q. a — o Li II V CO OO LJ < zz 0C s — LU a. z O — i. < «o □ c; -J— o CC LL LL _i LU r\icnr\jLPir**Lr*«»rMr\0OCT — ooc r^NOc^-i-ir vocxoooco— <\j inLPLOLCLOiOLr\ir>LoiriLnif\yc1CLO»oif\oniri/>4'<» .j-fnr\j— -■ ir^u^u"^inifiir>rvLfiinir.tf>u"\iriiriLr» O**G'C'0x0vCr0v0>0NO^C CTCT0vO0^CT'CTv0CsCTs0>G'C'0*,O0,0> OOOCOOOCCOOGOOODOOOOOCOCODCiO Lop>j*troove"-ror.j-r«-— (*iLrr*-v0f"-O<\l>$-r^.CCT-— •t^^CT — f^r^*OLO^r\jo^Orno^r^rMr^rvj(^f\JOCrooOroco<^ov'tO>«$-Ovtr _i^w^H-Jorocirir^'»(nmMNrt-i (V (\i f\J (\1 (VJ (NJ (M fV f\J — — —> — — I —t — — — « — < — —J — — — — — < — I — COOOCCOOOOOC COOtJOUOOOtUCOOO ^mccCT-^af^^^-LrorMcr-o^orricooir-JsCiC'iNiooccr-- r^r^rnfffVjrvj-jf-ii-iOOOCToor-^OvC-J-f^ — CTorvO^rvjco-CTT' ^^^Nt^Nt^^~T^~t^r'lrnr^r^rCrr,rrirArMrMrMfvl0^0AO^O^O^C^O~O^C?'C7'CJ'0>O^C7vrTOsO>OxO^O^ COCOOOOOCOOOOOOOOOOCOOOOOOC O or — LUO OLL au. (Nicnj— socr^crcsirvjvoo— "— '—Loirvoc^r^tr-oo-ocrar^o mL0-JL0croopltr>occc>C3>a>r^»CLr1rri(Nio^)>*'O^'NJ00vj-o r- co tr cr a- o o V- 1> a> ao f>- >o m >t (n co — o a> r*- -o ir\ fo <\j o or- cc j3-o^>-Lisrr«-r--«c>L!^^).c.£>>Ov0»CN0>ON0inLOLrim(fMnLr\>}--j OOUOOOOO(.iUOUOOOboOCUOOOt)C.OljL C'OOCOOOOCOOOOOOCCOOC oooc occ o Jf. o _ILU o «0 — < i±U 1— LL — CO >LL CLL TZ. LU u t— u or. LL' CC 3U co- co t- LU«3 3TQT a Ol- JJof\i ir\oo<~Lf\r^r\novn-o,f^-y<;mtT'p>o-o.^-or',r~~rr-cc— jci-j- a> ao co x r» i— Is- to vD o en ir\ -$• -j- m fn (\j — — o o> co r- io en -j- m — cococ^cccccocotra co coco ao coco ppcocox on r»-r— h-Kr- r-r- r<- ou'v;o'jccoqcoooopoc>nntioopoooo lo — yr (MCLxr^i^-orOvor~cc — co 3-r\jf\j^--c-4->ocoocrci'"ri-r — ocrroirrri— ^c — in— lnpl^^ovxJ■c^r*^x<"v3oc\lr--^\J^o oor:^o~c^o^o~^^'Xa^^^r~^fl^Qunir>r<^rnr\Jr\)— — doo^o^co r\jr\j — — — — -*-i— J—— *4ii—<—j-*—rH— ————*-••— ic?00 aoooooobodooooopooooooopoooo occ occ ofc oocoouoccacoooopconu uOul OwuDuO'JC'CuotiCouuo^'Jtic j^'C 1A1) ^o^oiAD^Q^o^OJitinoinoifo^Cirioiro — og rsj m r^ »}• vj- ir> lc o -o r» ~ co co ^ CJ" 'O < J — — f\j r\J f> ro >j- sj- lo ■a en 177 u c Cu C I cr> a. • •— rr < 1— I-O I CC n-u. c > < c c^c oD o f\i ii" "^ >c f^1 o^ cr ^c f^ (van c (sjr^f— '<\irv .^iPr^r-cocoffffffffooffffr^r^i/^.^r".^CO*C-l--*»Crr:l? ■* ropfirOffM^rfrofrifrf^^^tr-, fr., fr.fr, r^ror<"ifo<_ OOCOC OOOOOOOOOC DOOOOOCOOOOOO o ■z «3 s er c u. 1 c a. r o' 13 r\,rjr\j--ca.^fr1Cr-Cf<-.0 •*ff~j-ff>»ff^a'4-ff^ff>j-o,«i-0' c\> r\i r\ rg (\1 <\J r\i f\l — — < -- -- — < — 1 — — 1 -J — -H ^-4 ^H — ' — < — • — r-4 — i -j COLt-H.oCOtOi.Cl. '-IC t.C CwCC<_CCOCO Co '.p _ i — -four >r crc-a c r~(\jir. a- c >j---rr»-r\jcoa r\irr.if>innrir -f iP IT. -J- ro f\ if- ' O f— IT r^ c r- ff O «C f\ U". O -J' *— rviN]fNjrv'Mr\j(N.fvjr^rsj'%rvr\if\ — i.j—1— *,^oo Off 0"tio:r--c crff'ffffffffffffaffffcrffrTcrffffC'CT-CT'ffCTocepeca.cca.. coorooc c cc ooooooc coc oooooooc r.rHCT-OTO-T -CC— iNjrOf^O'^'Njr-r^cOO e,-4--Os*fv^crlC.r,~ X'i.ioiroooc'5 tnniicoiKi-iCiirm^iT-cr.c r- <\j x ^r cr r-jooatrooo j'J-Xir-vCtnNTfirM.-Hoaor— nOl< f rNiOtrt- uaooocotjooooooooooooooocooou oo ~->iiC eoc-ooooooooooi". >OOC CjOOOC' o p o o 1 J- i — a; on a a s> u cl s. x en a_> to cc cc x cc o r— r— r— r— r— r- r- r- r— ^o «o -.•< c '.C'OO'Jr >t: it'Onrpc t I 1^) — i at 3. r~ O r\j r\i CT o f^ C -C O r\j *r tf"i - "i r*" f<"i oi_>oooooooo->oooopooo-> .- 141 «. L i X UJ £ at a. '->- JJOJCl -'OOOOOO - _>w>0 _) ^ _>0.^ _IOO 178 IT C a «» • u_ C T or < o. • > k> 1— 1 ro U <: UL (\J luz a UJ Oixj a II «•— -i kU _j a. UJ 1/5 — < O -J 1 U. a: < o mil. UJ a «^ UJ > U. < o > - 00 C(J 'OJ < ac— • UJO Jt — L.U. au. e->ocr cood^ — 'f^-'CT,*l"r,"1iriir'-»- ko>c—MC1'<'>t'i''J'>t>),iTiWMO^ooiri-(,ioiriO>t(j r^MTT, ro^rrifnmfnroror''f-'»-'C>cj(7'oo C"C7NCM>CTC',0,G*O>i-to-J-<— — 'f^ sCsClo — ■ —' o o cc k- -J- cm or ir> f\i k- tt\ cc m od m oo r0l/Mn^'tC,''lNN^>-iO ■X— icrr<-ifV(\Jk-v}TM(Nj-j-i-i^0-LOpn(Mk-CD(\jrri— lO-f— 'vOf-a — j__irtrV]r\l<\jf\j<\J<\).-J.— i — D-CT'CTva,'CJ-coora:eo'Dor)anoo ' oooooc poaocoooboaoo^oocoouo xi^o'-r^coO'Looajcoccoor-vrLri^— rrio>o^:X-i"OsC^ONOvCiriLCMnLrir. irio *<• JOOOO( JODUOIjOC UOOOJOOUGLjOOO c • jfoo'- • uocoxoouccoonococ ooc c j. ct < — a -3U UJ't v" k- •-< tt y -re or— — .k-j^xTnvjcrx inco^oc^kCMM^j-rt — t<". -c -\j — i— lr\JLfMrMNJlNjrri-rvr-ra cr IMMHC? j> >r r*> m o x i; c\. j> >n rsi r- —. jj — ajDa,k-k-^vO-i.irMTv»T>rr00 0 opt MOJODoaoocoopoouoooooooor UJ £ U.L.' riico <- [ loocpo1 jbodooc ic of .ojLoC jwujIJOU lyouLuuu J .r- u?o^t)ifidi^oi^oK\pinqioo*oi^l mai^ j -■^•M-rm^ rtr L0>0-0'^k-CDQ0t7vC7sO'J-^- 1— *- f»- u < a. (\j U-Z Q UJ 55 UJ a. li «*- -i t-O -i UJ UJ — ■a <£ -J 1 u. oc UJ <. o > < cc D • UJ NM UJ o 1- u o < z - o z. !/• <1U OC (-or n Jt. u-O a. B 1 un * o ►- H* on UJ c < li- ti- J- t- ar • l/> a nt: ■-4 3 1/1 II > l/l oo JJ -(\j — iririirtn^i/itniriiPiniririifi^iriifNiriiriiriri^ij^u-' 0* 0*1 0 0>0vD0>(J,ON0*,ONO>Os0*O*G**t7'0^0*'O'O,O'Os c oooc eooooooooooooocooo f\jf*-oki^rt">to(NJir*o>ovccj'Ovs0^r^O'0TW®ii^>c^o ----Oi^O0-0,t?'0000f^r^-C>0ifMr\>rsJ-r^rrif\.'(v^-i (Nj <\i r\j rvi rv —*r* >-» —> <-t #-■■-•—< — i — ■—.-< — i — — ■ — — i — < ULLC OOOC oooooc oooooooc o frir'irf,rv(Vj-*.-iOOO,cocef— l^''nr\lOa^r-u■\rva'r- W O* 0 C* (T 0 Ox0vO,N0*0s0'01'0s0*u**Ox0s0s0'0'O**O' ococ cooooaoooocr-oooocoo on o o < u UJ < _l 3 uO c^'in^it -oir\f\jf--i-o^rrim— <^i"~r\j0-0^f"-r^rvic~ ir l.'CUGCOOOOUOOOC'OOOOQOOOU oooc cooDoaoaoooocoooooo & o £ o -p 3 >- o -JZ _J JJ < — xo — jo— -*r^crr^cr -r ^ cu ^ 0s ~l f^i inj -r ir< — u^— *r- >r jxi l»-C a r-(M>-^tifv'MfVT)(Vt-'iAfMvOaofv-n jj ar^r^i^r-,c^irmuTd'^J"',,MfvJ,-'O0Mi:r^.Ctf, a. Oil— OJ.nr~-c.r T ."*>?■ Ojcc-H — — O0* ^ifmmminjTO ■Ofn— JsXTiODr-K-C«OU^ir. -tmror\l(\l-^-4tiOCM7' — _,_,„! — _.,-. ,_._,_( — _<„*— — ,-i_i_ l— -. — OO O lOO'.i )OC'OOOOOOO0OOOOOOCJ a 'J OJ— I lit juoooo1 'U'j' >■ . . '.'(. ot loqoodioi > -^■Jvi/Oi. _*00_IO ,>i^Ol_ O O <_> O o O >_> <-/ i^ oiTi >x. i -iroir*oir. <>*/\i_".noirioir,f*^ir o^* — ,-X'COU*0'OO — <— <(M(Ni 180 IT c a ^ LU C I oc h- »— r>- u < CL (NJ LUZ a. LU CLU 0. II <■— _i l-O -1 LU UJ l/)>— LU < C > < ae O ♦ LU ** LU o K u o «a z r> u- L_ 1 LO a' r 1- re to LU c < LU Q. vt 1- or • l/~ a. a: r— ' O t/i II >- i/l CO UJ < zz ¥ 2 •-LLI 0. 5 Q»- >- < m lo L^ >1- PA ^- CD — i oo rr> ^- v0 f^ oo o* c o *-" — — < —> — ' o o o" oc r- lo r^ o o~ la po cc -a- o mpo>j-^^>d->t«l'>* 0'CTNCJNC'O,vCTsrjsoN(7*cy*o,soscy,o>cTxo,s ocooooooooooooocoooccoo r\i c rf s- P> — < r- 0-^A->£)LP>-4-vJ-OP--Oi/\P\J<\iir vtP\J oo-OLOPv(jN>CPvjCCpr,corr>arrP\cof\jr*-<\ir--<\ir--'-«>£*-' OOOO0>CT<0,'0CXr-r~-0sCLrMn^«J'r«1pr>(\|C\j,-i,— l PM (\l fV f\j — <— 1^ .-I r-l^rt —I — < .-J r-< r-l -* -* —* — " *-< ^H .-I CCuOCClC/LUOOOOCCCCLCOOC -jtmct-oc — iapp>ccT>r\jr-pp>^LO»j-r^rvjo-^-Lr>o-oc lt. vONcr-r^.OLOLTi^'rsji— iOp-loi— cco-j--^>0(\jloc/'PO rvjPMr\jrMrvr^PMfsir\ifMPvJ-j— <^-i— lOOor/acrr-r- u-c^crcro^a^c^p^o^o^a^cro^crcrcrao^ooocaDcooc oooc oc oooococoonoocooco r-ac vi-NrPvjo«*-pMoo-<— io3Ai(\jt5>oiA(Moop-Lr'.a — • T JTir poij>i-in-OP-r-^;i)sOvU-c-OsO>c~o-£ivC>OLP,'LOLr\jOLnLr\Lr.Lr4-^r^- t rJo•.->•. >coooooaoooeoc~iooooC""'0 o o 9 o n PO o o o -p o JO. o LL [\jajaPvjiTirop-f-'cc>arococJr-rsi>j"Lro^ON— >c~*— ^^'^rAr^tf-cP-cor^rT^oh-y^mi^xiiT'^Pvi vOO>c-j:ir>irinQ-j-p'if\jr\jr-oofor- -oirx-Pv— io- x o > a,- x a> cr> x to oc cci co co or cr to f- P- P-^ Is- P~ A- -C > < if . ' j " J 1 3 < ) ' infJOC >(_JC>' >< JOOfi ' >r ' En l/)t— LU-i)!\nj'ir>a>aur«-LP\.j'r-inrr\_<(xji>o>aDir>. rMcrauuN— ir--reofAaocMp--« .cxjir a^cir-PMmo^r c^aj-ncoa r-r— o-olpilo-t ^■rripriKj.'*— iooiM"* oaoooocjpoaooooobt'oooooo l i . s. . . i_ _ juoqoooocpi jooca mi(jj"jnoiPiOuiOi/i )LOpir>oiriOirM lti — ■r^r\iprMnj-LOO>OP~r»--cao(rooLj — •— irjcsi •rr>a--j->T-f-3--»->T>rLPMriLOLfMr\ir> 181 O LU > I X < a. • V h" *- h- t_) < K a LU > < ^ *'LO oc.l. cocoocaoooooooooccoo vCLO'T'-'acin^-i^jrvr-rV'r-rvjr-— crrcc'--'^vC^LrLP«*-»rn"\rrr\j"-4'--!c r\jr\j(Vrvi.-<.— -«»-l^.-l.- i-H^^H^r-irt--<-<^-<— —l utuCLLuCCOL LCL-UCCCJCLOL't' irr-irxLr>(\,r^o-r~oiNcrre,j--Hco^^rnr-xcrrM CTO^^-— i^oCrror--Lriro— xxr- ir^-r-.r~NC>j-rnoh-LrrMa-ipc»o^-' «JnO<; OsHriir.j'rnrNj—iLjcrror-LOj-ri^oCM^^ 4. vO O vC >C >u ^) ^0 sD O sO ^O LO IT LO LO U" LfMTl LC. >t ^J" ^T LJ' ^ l C OCCCOLCOnOLCOOUt'l o ' r)_)C oc oooaoooooooooocoo i — r\t,'~"— ■■"<-^j3r-jowMirooO'0,*o,o <— ■ o-t' £>(. x ototr^r^air>-orrr\ocD U ^.L",,j~Lr,Lrvj.sffri',>rr— il.'— '■ t; -rx 3. x Tsicxx/u; •oo i/II- r-n:^"..rir>o^>raor-NO-t'^Lr>t\jcrr-~r— *rn^*r- r > oo A)-ro-urrir-<\jr—**>oOLncrrrir i^^^TCMnj traiojX.xr-r-io-OLnm^^mrNjrsj— "^ooLrxtJ -<— — — i~<-j— i»-i— .— <— i— .-.— i^[--i-<-h— .— icoo! r;oooooooooobooooOP3l r Jji LLLJ OJ_J U'J'.'UUUUUUUOUOUOC'OOUJO-. -' --'-_/ .L/L JUJI . O ^ '_- l_* — I <_/ L_ A_> ^ -I r».jj~ ^ n^LOOLnoiTiOi^OLC.C-> X- jino.r> . >-p — J-Njr\jrr\rri^.^.jfNin^jx)r-r^XiXlt^LrOO'-<— H *- *-* (J U-' <^ o > - fir LU o <■ Uj f^cocooocMroopcooooooooaor^f- u~iLomLoirMTUMriLntOLOifMsCT*CTON^0*O,C7sCTC*t^ONk^0* ooc oooceooooocc c sr"^(v— i^-H-j-i/Mrcocovj-r^oLorv.! (K/ffC 4-c inoknokC OkOO^tC* rt p-ooc(rcr^r->OvOi^Lna~— *c vrm*j*r<\iOf*-LJ>i_ vCLOirirM^mLOLrLnLT^'*T t uoc faooooooop oo< looobooooocob o «H o -p 3 _ joj a. • j if jM-d tr CO fn -J- o Is- UO if\ >C r-- tT> •i — 'j- !• i — i cr (v x in*- (»■ . o* x r^ o ^3" r>- »-< • ' )PC ^C pi >oooooo On i/)>-« OOI— Uj« ci a: a - ijOl/u iknoinoiriOLnaifvoinoiop — "NjtMKimyr^-LOjrv^-i>p-r-xxt7^ 183 m c a it c I- «1 a > u c > LL •O a > < ty » U Ul (.■> u O 2 «1 «9 h- Z V. a 2 ec a. i_ > o OLL o fs. a it -* >o r- — vO •"ir a v* OOC.OO<_'<_>(_/OOOOOOOC_- o o d ii c c 1/) n 00 P-U < y a 5 ► •Jo — a. o— •oo uipicst-j-fvjccrr— ir\-*comoir,o OOOC OOODCOOOOCCO CvJ * — o uooocoac uoooouoo OOOOVOOpOOOOOCOD §1 o o 1 8 -J-L- « — a — >u. T J. a. ini- a. i> C7* O cr u > t\j xi t\i r- m r— r— i»« -u sp in ir\ ouooooo _> ao c.) in t\i <4) f\i -r MJOUOOO 3 O ■H d. <1> H ■8 _> i U.O ir>Otf>0 r\otr\Oir\ oinoiAOiro — 'rs/rvjrorn>j->rUMri~0>Of^r*"0DXi|7> 184 ir c O 4 • <£ II O I- lb o X co • H i— i f-4 o < a CO LLZ O LU OIL1 a II <•— _j t-O _i LU LiJ (/■)«-< LL < O > < or C » LL *— LL c K o o < 2 •c LUcr" <~ 1- O i V. • (/) 'JO U_ < Z7 TC 5: M-U. a 5. c »— • 2 >or^cir'*fvi-tooo»-' rcmi,<.rr.<'«icrimfri(nr«"i(\i<\ic\jrg-<.-< • ••••••• • •••»• •!• ceooccoooocoocoo an -o cc no c\j r^ C C °° O in r\i r- mm cr •J- — l-oin-*'CO-0Oin0>>t0Df\j l- C-J<_'_~<_;oOOOOOOC OOD tr ■*— >>t-'orin.-c'CO'x<3LOpncr--i-oiri — ireo crocj osovcrovc~0Ncr>cro",ooaoap«: OOCC OOOCOOOOOC OO r->orvnoirvcoer>.- c r- in ac — < m. ■*^,<*->TfMCJ,p»(*'-o«Lo (\j — ( r-« O O C7" 00 r» •£ IT. J" <\J — I CJ oc t- ■V O -L >C -O IT . m in ITi IT. IT .JIUMP vT vT idoocoofcoooocjc op oaocuocoouoocooo 00 o o O p o £ o ^3r _JLU ■>LL. LU L a— iT^-fO's-oxHai^^o* L'Mnc.TmnmMH^ < o~ a* r» vo in r" TTi(i'Grx'oci)trxtx>i^rr»r-r~f^ c idocjcopodooooop LO>- LOI— u_uooc oo .nam* iir air pincmouliOiAp — <<\|c\im-n.r>}-tf"\in.u-br-r»crcolr 185 IT O u o -I <1 a. ID c I 00 ■a Q. • >- K In IT U < a. 1*1 U.Z O UJ (JUJ a II <"- _J »-U -J Uj UJ l/)«— < o _l 1 u. or < c f UJ < a > - O 13 - c I i/i UJ c < a' a ~r >- t* • 0" Q. 0C »H c uo II V (/) OU Oj < zz a 7 •-.UJ a s a«-> 2. r»-omin#no OOfM^r^oOfOCirO oor^r->o-cinr>*~t — < — «~<, — I.— < — ■— i— -i — > c oooooooo — i— 1nDU~sOinpuMrvmir>^r • •••••••« joocoooco O o o £ O ! a Q 13 < — a u JLJ LUc>oir>4-«in o^iao^o-fcroooi-— i xcpcuxaoaccoic oo ■j-'Mccir.r-cro^i'-'^i r^oorri(7>f«aorrMi-(M vO^in^-^oim^iM — . _._,___._,__, ■•JL: CM H A 0) _)l- A- t XI s J.U UJ_I *u. OOUUOU -1 _> J • S< O l/\ O T> 186 to •4- II o ui > e u OJ2 OLD <—• UJ I/) oo LU c O _! — co o* >c co r^ r-- »— 'Croo • •••••••• coocooooo f^r~vrorOLrnr>Moc> NO(\ioo>j-a>*-o>'*oc OO0C^-t<\jCT'ir\<\l uoooOcofco O O C\J J- UJ LUO JL~ CLL au. — 'fM^l-f-.- -j — 1 0"> r- OLO>J-rr)r<-1<\J.-*CMXl irtriLnirnripinsr^ c >< iuoooouo odooc < jooc o -p p- ■p 4_J <~ -JUJ «3— • o m'J* -o to o -r ±-i -r i urad'o: xmi~- • «•••••{•• LjQOOC oooo -3LJ c/> — □CQC a LT~TOJTlC'U\OCnOU JOUOOOOpO JJ s 187 lf\ o o> *• • •o II c •" «r a UJ c r CO <■ CL • V ►- ►— IT o < a c~ U.if Q UJ t5UJ a II < — _J H-O —i Lb LU «/!"- •a o _i 1 U. tX «3 O C*U- UJ en z UJ > UJ < o > «T Of a » Uj *- UJ C t- o O <* z- < ujar - n-o» LU o < LL Q ~J- t- * • 1/ a. QC w4 □ 1/1 II > l/> <_rc_; UJ < 77 a. s M-uj a. z O — s «3 «o n u -u«— a rOU. <: o 7 X UJ ^: • t — UJ o m LUO ^ (_> -J- Jt — <3. in Cu. — 1 * II a j. 3 • L. 1 ^ .j- >r m -« a p- r> cm < at LL _u' o -JUJ U "3«- < 01 o ~r-r — ir\ < >o <\i coip-r»- — j_ c n(.T'3l.:OpO JO act. a u"v-r— \(j#Mcu ocrocr-Toccioofj-o -JOOOOOOCO CJ1- -**<. OOOO _'VOt o ■X u. r -o^ «•••••«•• a o r-l 'VJ <\J rr( r^ ,J- ,J- lf\it\ UJ-J ■T -T -f -T ~T -f -T ff -T H 188 IP o U C -J LL > -J < CL LU o T or <3 a • t- *~ a < QL C. a LL. a. n _j — i LL u. < O -j c^ «J 15 LU a Z > LU <3 nr m- LU LU C l_! (J Z» < <3 K :r LO ar a DC LL c a: c 1- LU c s CL st v~ • CO o: t—t c i/i II i/> LL - u LUZ OLU «>- vC KO f- iT,<— o-- 1 u. OD f^LL • LUa c m /> o pr-o» • a> ^* Oi_> ►-LU H LT\ :il~ cr- O O U_j — >LL LU ■3 EH JU l0»- JtCfc a. ■1 u C LLC 189 o a- • ii c y c X r<^ < a • >• 1- *- <* u « a. >* LbZ Q Lb IT Lb a II Lb < c > o or o t- fvO* • LU c < Lb ^ 0. >r I— Of • l/l c 3C p^ c to ii > in ou UJ «s .^z *-> ac £ —.Lb 00 a. 2 3*- r^ * <3 •«<_) X a O _)'— • u ctll o c 3 CM M3 ro O O 3 U — a (_ Lb aib o £ O I I -J7 _iLb ~-t ■a — co aec O a — a > u. • | O-b ' > 4) 3 or LT. — • C'l- XJ<3 -I hX t b < LLU Lb_l ncu. 190 o o o < a. t i — » — i o a. o • ■ * u — — . ~ x in in in < z x lu — a a. o u o z _J _J XI -J oo oo O < < < a. — » ► ► a: ~ ~ ~ o ~> rm -t >. oj »- Q O o < a. a. z > o IJJ OJ _J LU o 3 oo x o o a t- in j"n in »n o ■* in ■o -o l\J IN O O o o o o o o o o o < z •> LU XI J < «; > iu ai Q 3 nj 3 CC _l 2 O 2" Z — a. 3 3 o -' — w X 3 a -k a. a. a o x _)_!•- Jl oO —■ ,0 j a -X «■* • - I — « -~ > a. o> if» U ^ — * — * < — — 3C < 3 x 3 o a. a u < a j u z _i _i < on oo 2 < Q. LU ""■ ► » X -»■—>— 00 m r\ lu — « — I 00 at — — -*. 3 I >- JO a < oj — 3 >lu. J JJ J /I Q LU o, a. x x » » < h- -x iri'i> ji- -• — -J z — — -z j x 2 a, -o o »© h» go ^ s0 -o o o N N(M M o o o o o o o o o o o o o o o o o o o o ♦ » IN < 0 » LU O O I rn ui o> ~< CM O . *■ o >0 1 • oo O ♦ I 00 < ♦ ♦ -< < o * I <*■ UJ O m i 1*1 O INJ ■1 0> • f- O IN + • oO O » ♦ —I oO o » I OO LU » O - <\J IN o o o o o o o o o o a 'J l/> > o iTi 00 ■*• * CI oO • * XI o lO -< ♦ < it — xl tr\ 00 rr\ -. ( « o (M < r>- oo c* h- r«- (*- in in in o o o o o o o o o o o o o o o 00 fl • *-• O -< + f\l 00 • » o < + » o -r ct> o CO I — I XI — f- • O O t\J 1 -0 O »* (M 00 00 00 MMM o o o o o o o o o o o o 0 o o * ♦ < oo * * < 00 * * «»■ < o * 1 < LU * — < t»- * •c ■c (M O nj I t LU O »M I r0 ^»■ OO 00 c\l <\J O O o o o o o o o o 00 *> 00 < « < * # «t < o o I O -0 + =0 00 0* * >* < lO » • < o ♦ I ^ < oo o ♦ * I < oo UJ « «• O < oo N *> * ■O r~ < ■O 3 » ■O I < I OJ » O _i 0 * .-M O < r- I it • LU -J- o -< O I ^ i oo »r OJ » f>- LTl OO ^n M * • 33 00 O r\l -n >r 00 00 O O O O o o o o o o o i *H UJ •* m in CO LO o MM < X a. _i < o z < a x Q 00 00 00 G* NNNIN o o o o o o o o o o o o o o o o 0 o o o <\J < l/> I * * O < 00 1 # # oo rt < # o # 00 I < # LU # 00 ny < # M « < 00 **■ » oo O - 0 OO OO 00 r>- # in »f 00 (\J — # • —i < O . * + O < oo # # # OO fO < # o * oo I < » OJ * < oo < # (Ni # ** in >t 1 O I LU • LU MOS (NJ | 00 oo oo c* -*• * f* *f < -4 • * • 0 o o * K* I < nJ • UJ # -* O GT> < in + so * s0 < (M in • * r- o 0 }■ » • r»- o^ oo o co •«■ * | — i in «■> ^> m «o v o «N (N o o o o o o o o o o N 06 O 0*0 0 o 0> 1 «\ 00 -^ * I oo o * I O 00 0> * <0 OO in * -I < < # —t IM •(- o 00 I * LU o in i*i ei 00 (N XI -O •0 rr\ • • LU Q O '-l IN o o o (^ H"N o o o o o o o o o o o o o o o o oO IN # 0* 00 0> # • 00 O # I < 00 « * < OO « + o I IN IN m tr\ o o o o o o o o o o m in IN IN in m o o o o o o o o O -i < » < * H < o o — ' < u. » o> m >f o r- I t UJ o <=o + m 00 in * (M OO (M * • < O * + < O0 » * m < o » I < LU » ■o < -< # **■ o - I • LU o o> 1 fl oo IN * 0> < r~ * • < o * I f| < O0 O » » | < ^1 xi * ■»• 0"> < 00 (M •» » Ifl ^ < -" O * ** I < • LU ♦ o r>- < + CO » < -i o * -o o < -1 I * • IX) -n o -D O I -i f\j n -r < X Q. -I < Q Z < > 10 Q Q X o u. oo UJ O O0 0L O in -o IN IN o o o o o o o o IN O « » • o o + <2 LU c z > -j 3 a. -Jj h- oo oo IN IN IN ft fl 1*1 o o o o o o o o o o o o o o ~> • 00 ►• ♦ # * 00 < «»• * * • 00 < ♦ # # OO 00 cr> * #0 00 < I * # UJ < in m * O rr\ < 1 IN « Uj CD < in fi * h» • <4> O I O -« oo | h> » LU t oo fO ♦ # >o oo < —> » # (N O0 < -< « * • < IN | * O oo - I * O LU < -< # IN < (N «* # a o < ^3 — # oo in TO » < -n — 3 < + I O < < I ■» # XI |%J < ■N O «• -T | ro CO lU O 4- •O I o >»■ XI • CO -j- I o h- uo -r -< * • no yi I - * —i -m ^n oo «- 1 jj * -' o L^ < z # I- xi 00 LU □ * CD < < — _ * »» — < o o # o o u u >-) ■0 h. X' 3 o o 3 O 3 o 3 3 191 INI O o o < a. -d CD 0> o O ■»• mIm|— lOOOOOOOOOOO— 'mIm!.--—!.-—!— i— ImImImi— I— ir\lj™<>j— i— i o o o o o o Ml f\. Ml f\j -4f\J — I M — < - < in j\ in it Z ii II II II < • » - » i 2 H- a. < 3 > _j j- < «5 T _ 3 _1 i/> /i < lO < z JJ < z a c z V) < i— < > >— jT < Z. a □ E < Q. m 00 m» mi mi m» w JJ — LU < JJ -~ -j i — z < z — UJ pm ►■ < t- ^ -j — j < i: — —I — . Q < a * M. ^ — . m. «m — . LL Q — UJ c •s & J. i/) *^ z Q ■s a. U_ Q M >- D N a. Ii. L0 t— i > a ^ a LL LO <~> a Q >v a. m o — O O a o < II Ml u II UJ II u ii — a. II z II a II m» < II h" It O II — a ii s: ll I II WJ a. II a II » J1 LO o o o o J — » UJ z M] a m. J — z ■_i — a ~ u i — i r- > — » Q — o — . a; u *^ a •• J — Q u — * a ^* 1M t~ r-- 3 n. i*- *~ r^ -* mi u jj Ml * — . — Ml X —j — ■ * M — • — < JJ — . # — . N— — . a. _i — i «■ Ml —• M] a. -J Ml ♦ Ml — » p * « •> » LL •m J. Ml 3 vi mi • mi a. mi > a — • M LL — i a 0^ mi • Ml LL — o v) — • M> j-i Ml -0 JJ Q J"> ,n IT x» II UJ t-*J 1 UJ o JJ II z — 1 1 z o Z II i — JJ i i— O K- II *- _i 1 i. 3 i- II a _l 1 3 a a Ml • M. Mf J — — ~ m» DO o JJ a o UJ a o O ^ a. i/l Z a. o a. t~~ < O ♦— < Q < c a v) i. a. O a J a. LT) Cj a a a. 1 —t \T\ a LU JJ > ."> 'J J a • • L3 z « •c z — i 2 a. a 1 < < > — i > a o JJ II ii UJ II LU a _j ii II -j II _i □ -J II II _i ii 3 MP *— < Ml Ml » _U JJ LU uu II II II LU o UJ Jj a jj a _ i l0 Z z i/i z vi jj O f— i— Q *— a j vi z: i. v> 5_ U0 _i IS) a a v) a j-i LL II II a CC Qt 'J TC -X. 3C at 1/5 < — 'J (J •a a Q Q. a. l0 u o- «(\im>rx>or~'uOM im«i^^^)moto- i r\) rn ^ jTi o ;_)— i— i— •— i— imimimimi— ir\j \j-\j \iMiM'>j'\iM(M|'liT;firnmm'^fnmi')^^1r-r^^-)"'T'<,'*i^-'>iJ>i' ■♦ in r- co O —I (NJ ^a* o ■# in r-oo o -* tsj >r m -o » _j _ — • _ I M — I — MMMMMM^MMM— <^^ OD j- < LU 3 IX >0 >- •s. > Q ►- -* « O (J Q » "-' fT| I • < LU X > «v h uiz h in uj >0 ■ Jff < »Q _, x O — i _i v — — — — «- Z • 3 U| — MM — — -O < X U LU w — — — — m X M _j a; _j LU Z t- SL a | •* ► <0< o a- < a a. a. — * lu o uu 3 z o > a o » o i a t- ai -j uj _i -j — x z r»— o o oo o oo ifl < v m LU M^»< o. a. o_ a a. i/> v ^ Q • oc lu X o - - - - » sO — — «■ — — Z • _J (J in 0£ O -< — • _i » — — . — — woz >oez • >» f«- mm w w w w Z U. Z M OC UJ < X -LU ■c lu z ►- r o m • < oxi-^> wo II Z O > O O WHmilJ MJZ-JZ JZ lu _J lu _l _JLU3h-LUt0< JILUKm Z i- 3 a Q o o_jt-a:_jt5ai'^t-o>o<»— •./> lo — i — — — — —i _| 3IUJ> 34 — — — — w _J_JLU<-»L0h»O_JLU O0 LU lu z h- z: 3<3i_>o>«.30"*ooa£> lj Q- < a a.-Ooozxa-o<»z jj 2 O > 3 Ot-XLU30o£Ix<3>3 LU _J 'XI _l J U rf Q N Of LU 1*1 ■« Ul 3 3 oo 3 o»» <-«••« *,a:» — i »li>- Tv O O O i_) U 'OXLU »Q3< —i «. •■ » ► ► lX,XZf\JZXLUo0_ja£3 — — — — — Orr*»»»>-0>-a£3C3 — — — — — • ^LUZac»-a: — — — — — zvr*:oLuaoo_i>_ii_>c> XI Z t- 5. Q0>»LU20H-LU_1<- j a < •x a. i/iMfiiuocuaiozo-Lnj Z 2 3 > "3 3 -Tiauj<4jDIS < — xi _i xi _i _j at a < -< au.o<"-oaLLua. I— MOtt. s: < < < <. rgujy>a — — — -. — i U>-LU-JX_IXI_JXI» X _l * — — — w w ioM- >- >— i— r~Q>rH-or^2i— (nj — xi-'-vj-^. XI X* XI jj XI *LUf^OXlr\J«i_)<\J*— OXmQ a co cc -a =q — xh-»J-XX.- luimx • ji"^M«iir\N^ps nfiuiN.jjf(iL/\r- • »ji^)j>o^O'!?>a » cr> o -ijijiinHji^-nm>ifl-iiA^^^ii\j-i3flH3«i »- flui ^fi \ « f— »- r— » r- r-- »r~"«-f~- »■ f>» r» f«- « r- ix x O ^ ► O 'v x »\3M l\a. 3 —!»►►-< »»_■»»»_ ■..►-^•.tn— «_i>vX_jvf>xv, — < — i v iT> ii 0 >0 O II -0-OH H O v il J O 4 II -0 — — < — LT\<— — rn«-_j »— xxx xix. -j_ix xi-jixi lu<< <m<^ LT>-or^ x o m la j> m j^ m 't\ lt •H ■P O o o 'j .j 'j rj o o JOOOOOCJOC300 OOO J 'J 'J 3 ', . J^uuUOOOOO -o, ,2409999467 -0.2518250346 0.0108250380 -4.4917144775 t J, . OOOOOOOOOO -0, ,2 /r,9499504 -0.2889920473 0.0119920969 -4.3292751312 c->, .ouOOOOOOOO -0 . , 32o9999o00 -0.331 1289549 0.0101289749 -3.1554431915 70, ,0000000000 -0, , 3759999b 71 -0.3787130713 0.0027130842 -0.7215649486 Mul ;i Ctoftclb ObVIATlbN VERSUS ANGLE ACionl ucVUTlL., CALCULATED DEVIATION ACTUAL - DEVIATIONf 0.0014475845 -0.0214747824 -0.0293298364 -0.0258062482 -0.0145923495 0.0006214976 0.0161475539 0.0282956958 0.0333766937 0.0277073383 0.0175876617 0.0093402863 0.0092658997 0.0136795044 0.0288963318 0 .0 0, ,0 -0, ,0014475845 -> .OOOOOOOOOO 0, ,0599999987 0. ,0814747810 i i .OOOOOOOOOO 0, , 1299999952 0. , 1593298316 1 5, .OOOOOOOOOO 0, , 2099999 7o5 0, ,2358062267 •OOOOOOOOOO 0, .2999999323 0, ,3145923018 . i .OOOOOOOOOO 0 , , 34499997(>2 0. , 3993784785 1 .OOOOOOOOOO 0, ,5099999905 0, ,4938524365 n .OOOOOCOOOO 0, ,6249944952 0. ,6017042994 1 1 . JOOuOOOOOG 0, , 7549994905 0. ,7266232967 .OCOCOOOOOO 0, . 3999999762 0, ,8722926378 >0 .OOOOOOOOOO 1 ,05494S45.000G0i.JG00 0.3750000000 0.3727575541 0.0022424459 0.5979855657 70. 0000000000 0.3679999518 0.3838108182 0.0041891336 L. 0796728134 SLOPE FACTOR DI/DD VERSUS ANGLE/ ANGLt, LLoKLLb ACTUAL FACTUR CALCULATED FACTOR ACTUAL - CALCULATED PERCENT ERROR 0.0 U. 0769999623 0.0754218698 0.0015780926 2.0494716552 b.OOOOOGOOOu 0.0789999962 0.0780347586 0.0009652376 1.2218189240 lO.OGOOGOOuOG 0.0809999704 0.0802936554 0.0007063150 0.8719941378 15.0000000000 0.0829999447 0.0826053023 0.0003946424 0.4754729667 20.0000000000 0.085999965/ 0.0853770375 0.0006229281 0.7243353128 25.G0000C0000 0.0H99999738 0.0890154839 0.0009844899 1.0938777924 30.0000000000 0.0939999819 0.0939274430 0.0000725389 0.0771690011 J5. 0000000000 0.1019999981 0.1005203128 0.0014796853 1.4506711960 40.0000000000 C. 1109999416 0.1092003584 0.0017995834 1.6212463379 45.0000000000 0.1209999919 0.1203768849 0.0006231070 0.5149644613 5G. 0000000000 0 . 1 31 9999t;95 0.1344547272 -0.0024547577 -1.8596649170 55.00000GJ0UO 0.1499999/62 0.1518397331 -0.0018397570 -1.2265043259 tjG.OOGGOOOGJO 0.1719999909 0.1729413271 -0.0009413362 -0.5472884774 65. 0000000000 0.1979999542 0.1981655955 -0.000 1656413 -0.08365 72051 /O.OOOOOOOOGC 0.2269999981 0.2279177904 -0.0009177923 -0.4043137431 Table D-17. Program CHECK, Listing and Output (cont.) 195 APPENDIX E PROPOSED DESIGN OF AN IMPROVED INLET DUCT The present inlet duct is deficient because of the shape of its bellmouth, because of the small velocities that occur in it during operation which make accurate determination of volume flow rates impossible, and because it has no provision to eliminate effects of atmospheric gusts or winds. The entrance bellmouth should have more gradual changes in wall curvature to eliminate local separations at the entrance to the cylindrical duct. It was stated earlier that the velocity head in this duct with 36 in. diameter is about 0.6 in. of water, a value too small for accurate measurements. To avoid large fluctuations of the read-out of the Texas Instruments pressure gauge, for small variation in flow rate, the velocity head should be increased to about 6 in. 2 3 of water or 33 lb/ft , at a volume flow rate of 360 ft /sec, giving a response at the instrument of about 8000 counts. For an assumed 2 incompressible flow with a mass density Q • of 0.002378 lb sec / 4 2 ft , and a velocity head q of 33 lb/ft , the velocity V. in the measuring plane of the inlet duct would have to be 2* %l[ V; = \1 n~. (ft/sec) (6) 2 Hence the required flow area is 360/166 =2.16 ft , corresponding to a circular area with 19.9 in. diameter. Because of the small pressure rise of the compressor it is necessary to convert the velocity head at the measuring plane into static pressure rise by arranging an optimum diffuser between this section and the inlet 196 pipe to the compressor that has a diameter of 36 in. Test data of Reference 11 show that the included diffuser angle should be 8 for the present application. Larger diffuser angles produce separation with associated flow instabilities, and smaller angles require ducts with increased lengths where the boundary layer growth is excessive and reduces pressure recovery. For a diffuser angle of S the length L of the diffuser is Z fen 4° Figure E-l shows the design of the proposed inlet duct. A well designed bellmouth from a diameter of 48 in. to the throat of 19.9 in. requires an axial length of about 26 in. , giving a total length of the duct of 116 + 26 = 142 in. Ahead of the bellmouth inlet a structure will be attached to support several layers of fine mesh screen to eliminate flow disturbances by gusts in the surrounding atmosphere. The whole duct is mounted on a trolley which can be rolled onto the apron outside the test cell to permit simple attach- ment to and removal from the presently installed cylindrical inlet duct. It is recommended also that an additional honeycomb flow straightener be installed at the entrance of this duct. The bellmouth and the diffuser could be molded out of plastics, reinforced by fiberglass, by using a wooden template which is split at the smallest diameter. 197 c o o. < O D Q LU -I O UJ > o CL Q UJ CO o CL o CL CL I LU e> 198 INITIAL DISTRIBUTION LIST Copies 1. Defense Documentation Center 20 Cameron Station Alexandria, Virginia 22314 2. Library 2 Naval Postgraduate School Monterey, California 93940 3. Commander, Naval Air Systems Command 1 Navy Department Washington, D. C 20360 4. Commander, Naval Ship Systems Command 1 Navy Department Washington, D. C 20360 5. Dr. 0. H. Johnson 1 Naval Air Systems Command (Code 330 B) Navy Department Washington, D. C 20360 6. Capt. A. Bodnaruk, USN 1 Naval Ship Systems Command (Code 6140) Navy Department Washington, D. C 20360 7. Office of Naval Research (Power Branch) 1 Attn: Mr. J. K. Patton, Jr. Navy Department Washington, D. C 20360 8. Chairman, Department of Aeronautics 3 Naval Postgraduate School Monterey, California 93940 9. Professor M. H. Vavra 3 Department of Aeronautics Naval Postgraduate School Monterey, California 93940 10. Lt. B. C Marshall 4 Air Antisubmarine Squadron 35 Fleet Post Office San Francisco, California 96601 199 UNCLASSIFIED Security Classification DOCUMENT CONTROL DATA - R&D (Security clsssUication ot title, body ot abmtract and indexing annotation must be entered whan the ovarall report ia classified) 1. ORIGIN A TIN G ACTIVITY (Corporate author) Naval Postgraduate School Monterey, California 93940 2«. REPORT SECURITY C L ASSIFICA TION UNCLASSIFIED 2 6 CROUP 3. REPORT TITLE EFFECT OF STATOR BLADE ORIENTATION ON THE PERFORMANCE OF AN AXIAL FLOW COMPRESSOR 4- DESCRIPTIVE NOTES (Type ot report and inclusive dataa) Thesis 5 AUTHOR^.) (Laat nana, ttrat naaia, tnlttml) Marshall, Bruce C., Lieutenant, U. S. Navy • REPORT DATE September 1967 7a. TOTAL NO. OP PASES 199 7b- NO. OP REPS 11 e< CONTRACT OR SRANT NO. b. PROJECT NO. 9a. ORIGINATOR'S REPORT NUMBERfSj d. [kK&xjLA^SM <$Zik 9b. OTHER REPORT HO(S) (Any other numbers (hat may be assisted this report) 10. AVAILABILITY/LIMITATION NOTICES :t controls and each transmittal to* e*%^m>mei^meai*ammmmi\ mini mr the Naval 11. SUPPLEMENTARY NOTES 12. SPONSORING MILITARY ACTIVITY Commander, Naval Air Systems Command Department of the Navy Washington, D. C. 20360 13. ABSTRACT A mean streamline analysis of the effect of stator blade orienta- tion on the performance of an axial flow compressor was performed by means of a computer progranio Measurements were made on a 3-stage axial flow compressor at the Naval Postgraduate School at six stator stagger angles between 23.8 and 44^3 for a fixed orientation of the rotor blades. Maximum efficiency and pressure ratio were measured at a o stator stagger angle of 31.8 „ Results at other blade settings showed that by varying stator stagger angle with flow rate optimum efficien- cies and pressure ratios can be achieved over a wide range of operating conditions. The results of the analysis were compared with the measured results, Suggestions are made for improving the manner of adapting cascade test data to performance predictions. By applying a non-dimensional deflection coefficient it could be shown that minimum work input corresponded to maximum efficiency. The test compressor has a tip diameter of 36 in. and a hub/tip ratio of 0.6. The blading tested is of the free-vortex type with a design degree of reaction of 0.5. Tip speed was about 185 ft/sec. DD FORM 1 JAN 64 1473 201 UNCLASSIFIED Security Classification UNCLASSIFIED Security Classification key wo R DS compressor compressor performance axial flow variable stator stator blade orientation DD ,T»M..1473 (BAao S/N 0101-807-682 1 202 UNCLASSIFIED Security Classification